Midwest States Pooled Fund Program Consulting Quarterly Summary

Midwest Roadside Safety Facility

07-01-2015 to 10-01-2015

what is the purpose of beam guard in the photo attached

State: WI
Date: 07-01-2015

The RDG indicates that the guardrail shown on the attached file should be used near a short radius system.  Any idea on why?

Attachment: https://mwrsf-qa.unl.edu/attachments/53690234fd13a4c0fbf7cfe65ff5dbb9.docx

Date: 07-02-2015
I would assume that the guardrail shown upstream of the short-radius system in the attached figure is based on two related concerns.
1. All previously developed and tested short-radius systems have focus only on impact locations ranging from the center of the nose of the system and towards the primary roadway. Thus, the behavior of the system farther up on the secondary roadway is unknown. 
2. Based on potential vehicle trajectories and runout length calculations, current short-radius systems may potentially be insufficient to shield the range of potential impacts.

Placement of the additional barrier  on the upstream side of the primary roadway limits the potential for these concerns. We have currently been working on a research project for NDOR to develop a treatment for intersecting roadways that better addresses the hazard area and has significantly higher capture capabilities. The Phase I report for that should be out this month and Phase II started July 1. 

Guardrail flare rate with Bullnose

State: NE
Date: 07-01-2015

With w-beam tested on a 7:1 taper rate, would you expect
thrie-beam to redirect a MASH vehicle properly on a 7:1 taper?


What about thrie-beam on a 8:1 placed in a 45 mph median?

This is attached to a bullnose.

Date: 07-06-2015

We have not recommended increased flare rates on thrie beam at this time. While the work done with the MGS system would indicate that the potential for increased flares with thrie beam exists, there are too many unknowns. The thrie beam system would have different capture, post stiffness, and dynamic deflections as compared to the MGS. As such, it is difficult to ensure that the performance would be similar. Certain design changes might need to be made regarding the post embedment, splice location, and the blockout geometry to ensure that it would work.


As far as the bullnose system goes, we have recommended standard flare rates as recommended by the RDG beginning at post no. 9 on the oncoming traffic side. Flaring prior to post no. 9 may affect the performance of the bullnose. On the reverse direction traffic side of the system, we have allowed more aggressive flares as the guardrail would be flaring away from oncoming traffic.


The RDG list 8:1 flares as acceptable for 40 mph and 10:1 for 45 mph.



Curb Offset Behind Guardrail

State: KS
Date: 07-07-2015

When you have time could you please give me a call to
discuss the attached? My calendar is full today, but I have some time tomorrow
from 11:00 am to 1:00 pm or anytime Thursday works well.

Attachment: https://mwrsf-qa.unl.edu/attachments/8d0fb26a6d0eca56a2fb1a16eeb4799f.pdf

Date: 07-09-2015

I reviewed the schematic you sent regarding a curb offset behind a guardrail system to help control erosion. You had asked about our thoughts on the offset of the curb behind the barrier and the height of the curb. We have several comments on this setup for your consideration.


1.       We would assume that this setup would be made with the MGS system rather than the previous metric height G4(1S) system. Thus, all of the remaining comments are made with respect to the MGS.

2.       As you noted, the guardrail posts will be installed in a 4" thick concrete pad with leave outs filled with grout. I assume that this is based on the TTI research on mow strips and leave outs. We would recommend using that research for the leave out details.

3.       The main concern with this type of system is the interaction of the vehicle with the curb during deflection of the guardrail system. For an MGS system, that deflection is typically in the 4' range for TL-3. We have seen that vehicle interaction with curbs can increase rail loads and increase the potential for vehicle instability. However, we believe that the curb can be placed in the later portion of the barrier deflection and still maintain safe performance as the overall barrier stiffness will be lower at that point in the impact.

4.       With regards to the barrier offset, the MGS with an 8" or a 12" blockout is approximately 18" to 22" deep, respectively. A 24" offset from the back of the post to the curb would make the offset to the curb 42"-46" depending on the blockout size. This would be at the latter region of the MGS deflection. Thus, placement of the curb at the 24" offset should have minimal effect on the barrier performance.

5.       As far as curb height, we believe that a 4" curb would be a more forgiving setup in terms of its effect on barrier performance. However, we believe a 6" curb will work given the 24" offset and the limited vehicle interaction with the curb.


Let me know if that answers your questions or if you need anything else.



top mounted Guardrail to culvert attachments

State: KS
Date: 07-22-2015

I've attached KDOT's current Standard Drawing (RD617E) for attaching MGS to low fill culverts as well as a sketch (Attachment to Low Fill Culverts) from a project currently being constructed. The design engineer on the project has run into an issue where the steel plate that rests on the ceiling of the box is conflicting with the fillet in the waterway opening (this is occurring at several locations). The sketch in the second attachment depicts the issue. We've investigated shifting the guardrail installation to avoid the conflict with the fillets, but unfortunately we are unable to avoid them given the needed post spacing. I'd like to discuss whether or not we could include washers between the steel plate and the culvert ceiling to shim the plate down an inch or so to avoid the conflict. I marked on the second attachment where the washers would be included. 
Attachment: https://mwrsf-qa.unl.edu/attachments/4dd0c01034511e5febc1e90c7dbc5297.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/dee23cc9cc5a36c8c28ab6ba30cb993d.pdf

Date: 07-22-2015

.  I do not see a problem with utilizing washers underneath the culvert top slab.  In fact, a similar attachment scheme was full-scale crash tested back with the original evaluation of this system – see report no. TRP-03-114-02 page 111 (available on MwRSF website/research hub).  The tested configuration utilized only rectangular plate washers in the attachment of the through bolts to the underside of the culver slab. You could use this concept too, if desired.  If you utilize washers of a similar size, you would not need to include the full-size washer plate in addition to the enlarged washers.  The individual washers alone would suffice. 

fall protection on parapets

State: WI
Date: 07-22-2015

We have had our contractor's mount hardware on top of our 32" bridge parapets to comply with OSHA fall requirements.


Here is one example. 

I have concerns from a roadside design perspective that:

I have heard of 2x4 being used  on our parapets and temporary barrier as well.


I wanted to get MwRSF's impression of what our contractors are doing.






Attachment: https://mwrsf-qa.unl.edu/attachments/beb3b76612f4a35076adcb4b0a121ff1.jpg

Attachment: https://mwrsf-qa.unl.edu/attachments/8e95b8975d4000173df1f12e99ff05fc.jpg

Attachment: https://mwrsf-qa.unl.edu/attachments/450f13ee326c40d54504e4b3f3edd5fe.jpg

Date: 09-03-2015
The concern with the addition of structures on existing barrier systems would be similar to those concerns with barrier performance raised in previous studies regarding pedestrian rails, combination bridge rails, and the Zone of Intrusion (ZOI) study. Additional structure on top of a barrier may pose a snag hazard, adversely affect vehicle stability, potentially create dangerous debris, and pose a hazard for partially ejected occupants or occupant heads among other concerns. 

Thus, while there is a need for fall protection in critical areas, we would recommend that the fall protection concern be weighed against concerns for the effect on the safety performance of the barrier. 

Development of safe fall protection hardware would likely require further research, design, and potential evaluation testing. additionally.  the design of the fall protection may differ significantly depending on the type of barrier it is used on.

Tie-Down Anchor Bolt - Length Change

State: MO
Date: 07-29-2015

Ivan Schmidt at MoDOT has raised a question regarding the
anchor bolt length that was used to tie-down the steel straps. The built-up
strap material at the hole locations was ½ in. thick. In the original report, a
57-mm (2¼-in.) long bolt was used. Original report link is provided below.

Original Report


In the letter report to NDOR, the CAT anchorage testing
program utilized a 1¾-in. long bolt with the RedHead Drop-In anchors. Letter
report link is provided below.

Letter Report


Do you know why the hex head bolts were shortened by ½ in.
and did not use the original bolt length of 2¼ in.? I have quickly looked
through the original and letter reports as well as current RedHead anchor
information. I was unable to find the published threaded distance within the
anchor cavity. Let me know if you recall as to why this both length was
reduced. Thanks!

Date: 07-29-2015
The bolt was shortened based on concerns that the longer bolt could bottom out in the drop in and cause it to break the tabs that expand at the base. 

Date: 07-30-2015

So, then why was the original report not revised? Also, the 2007 letter reports that the ultimate tensile capacity 17.3 k of the red head anchor is based on “limited testing and review of manufacturer test data of the drop in anchor conducted during the development of the steel strap." Seems like the tensile capacity would be a pretty big deal since these anchors are pulled out of the holes in barriers directly impacted.


And, if the tensile strength is actually based on a 2 1/4" long bolt, then the alternatives would be overdesigned.


Where and when was this change in bolt length documented so we can have for our records and should MoDOT immediately change the 2 ¼" long grade 5 bolt to 1 ¾"?  If the 2¼" long bolt was crash tested and worked why change to 1¾"?

Date: 07-31-2015

See my comments below!


So, then why was the original report not revised?

**There was not a problem with the original configuration. The RedHead anchors were used with 2¼" long bolts, and the barrier system performed well. The original report documents the successful crash test and installation details. No hardware problems were encountered in the crash test. As such, there was no consideration to prepare a revised version of the report. However and based on your inquiry, we are wondering whether a notice should be posted to better indicate this bolt length change.


Also, the 2007 letter reports that the ultimate tensile capacity 17.3 k of the red head anchor is based on “limited testing and review of manufacturer test data of the drop in anchor conducted during the development of the steel strap." Seems like the tensile capacity would be a pretty big deal since these anchors are pulled out of the holes in barriers directly impacted.

**Per Bob, the RedHead anchor socket has an internal threaded length of 1¼". Thus, the original longer bolt penetrated farther into the void region below where the internal threads ended within the socket. Those extra threads did not engage the socket and did not provide additional tensile capacity. The new 1¾" long bolt with full threads engaged the entire threaded region of the socket when considering the strap thickness and welded washer plate thickness. Thus, the tensile capacity of the anchor was not changed. The anchor socket behavior was controlled by concrete fracture and/or bond failure.


**If bolts are excessively long, there could be a potential for really long bolts to contact and rupture the deformed tabs at the bottom socket. We do not recall that scenario occurring in our actual field installation that was used in the crash test.


**With additional information from the manufacturer, we chose to use a 1¾" bolt length in the follow-on study that evaluated alternative mechanical anchor hardware. Again, the tensile capacity of RedHead socket was controlled by concrete strength. Both bolt lengths would provide equivalent tensile capacities when considering a 1¼" threaded length within the socket.


And, if the tensile strength is actually based on a 2 1/4" long bolt, then the alternatives would be overdesigned.

**See comments above.


Where and when was this change in bolt length documented so we can have for our records and should MoDOT immediately change the 2 ¼" long grade 5 bolt to 1 ¾"?  If the 2¼" long bolt was crash tested and worked why change to 1¾"?

**The 2007 letter report documented this change. Drawings were also prepared for AASHTO Task Force 13 Roadside Barrier Hardware Guide. The latest drawings are attached. Note that the anchor socket did not change. The 2¼" long bolt met crash testing requirements. The 1¾" long bolt will also meet crash testing standards as all bolt threads are engaged in 100% of socket threads. The shorter bolt length can reduce any potential concerns if an excessively long bolt would contact deformed tabs and cause damage. Although we did not have that occur, we took advise and reduce the risk of it occurring.


**It does appear that the original 2003 FHWA eligibility letter, B-112, still depicts the 2¼" long bolt for use in the socketed anchors.


**Alternatively, the online AASHTO TF13 Hardware Guide provides a 2007 SWC10 detail (8/31/2007) that shows the 1¾" bolt length. The link is provided below. However, the latest version of the SWC10 drawing that is located on our internal server is dated, 10-22-2008. Thus, the different dates causes me to raise the question as to why the online hardware does not depict the most current version of the detail that remains on our server as it should. As such, I need to speak to my colleagues to better understand how often revised details are forwarded to TF13 for replacement in the online guide.



**Please let me know if you have further questions regarding my responses to the items noted above. Thanks!

Temporary Barrier on Bridge 9123

State: MN
Date: 08-04-2015

I need your help with this one (See specifics below).


We use the Iowa F-Shaped Temporary Concrete Barrier System
(SWC09), and we do have the Tie-Down Strap System for the F-Shape Concrete
Barrier in our standards (SWC10).


We essentially will have TL-2 conditions (45
mph Posted Speed). 


Is there a configuration/location that would
not require anchorage of the barrier for these TL-2 conditions.  I realize
that a barrier pushing up against the curb would easily tip over as the curb
would act like a hinge, but how close is too close?


Is the top of the sidewalk an option at
all?  I can visualize barrier deflection followed by vehicle vaulting on
the curb (assume an 8" curb).


If we place a system with the Tie-Down Strap on
the shoulder,  How close to the curb could it be placed?  In the
Tie-Down test FHWA letter, I believe that I saw 12 “ (305mm) of dynamic
deflection for the area of impact, with a TL-3 test.  Could we go 6" with
TL-2 conditions?  We would like to give some room to the vehicles (the
barrier is 22.5" wide).



Of course the designers need an answer asap.


Please give me a call if you need more





I left you a voicemail, but I thought I would follow up with
a clearer view of what was needed. 



Some information about the site:

TH 21 over UPRR (Bridge 9123)

45 mph on the bridge

Expect that this condition will last for
approximately 5 years until the bridge is replaced



We are looking for guidance on the following questions:

Can we place temporary barrier on the shoulder
or on the sidewalk?

If so, how far from the edge of the sidewalk or
from the gutter line does it have to be?

What sort of anchoring is needed?

Any other guidance we need?


Attachment: https://mwrsf-qa.unl.edu/attachments/ec81acdbe36a98bfbf5393571fd8a16a.jpg

Date: 08-06-2015

I will try to provide my thoughts and be as brief as possible due to MnDOT's time constraints.


First, the duration of the work zone is noted to be five years, which is a long duration . Thus, a conservative approach may be justified using best available information.


Second, the placement of a single row of F-shape PCBs on top of the sidewalk may create some concerns. Although the F-shape PCBs will displace laterally, I wonder whether impacting passenger vehicles will reach a higher effective climb height on the barrier face relative to the lower bridge deck surface, thus potentially leading to slightly greater vehicle instabilities. Granted, I believe that this risk would be much reduced at TL-2 conditions as compared to TL-3 conditions. However, I just want to denote this low to moderate potential safety concern.


Third, the laterally-deflecting PCB on the sidewalk could create a scenario where a partially-redirected vehicle allows for a wheel to catch on the sidewalk, thus leading to increased roll or yaw behavior and increased vehicle instabilities. Although this behavior and risk is uncertain, I still need to raise this potential concern.


Fourth, a free-standing PCB at TL-2 of NCHRP 350 would result in reduced dynamic deflections as compared to TL-3. Unfortunately, we have not previously conducted testing nor simulations at the TL-2 conditions. Instead, MwRSF conducted a study to evaluate conditions at the 85% impact condition – 2000P pickup truck, 36 mph, and 27.1 degrees. Details of this R&D is provided using the link below. Note that the TL-2 impact condition provides approximately 35% more impact severity than the 85% impact condition used for the noted study. Thus, the anticipated free-standing deflection for F-shape PCBs would range between 24 and 43 in for NCHRP 350 TL-2 impact conditions. One might estimate between 32 and 36 in. for TL-2.




Next, I agree with you that free-standing PCBs on the bridge deck and adjacent to the sidewalk would laterally deflect backward and strike the sidewalk. Once the PCBs had bottomed out against the raised sidewalk edge, the PBCs would be prone to rotation without translation, which could increase vehicle climb and subsequent roll and pitch angles as well as vehicle instabilities. As such, adequate space would be necessary between a free-standing PCB and the raised sidewalk, even for TL-2 conditions.


Based on this information, I believe that a reduced-deflection PCB system may be worth considering for your application and located on the bridge deck. There exists the (1) tie-down strap system, (2) vertical through-bolt tie-down system, and (3) WisDOT steel tube and saddle system for use with the F-shape PCB system. All three systems alone, or in combination with one another, may provide a workable solution, especially where you have limited width for PCB placement. If you can accept deeper holes drilled into the bridge deck for either through-anchor bolts or rods epoxied into partial-depth holes within the deck, the through-bolt system offers a low-deflection system, especially at TL-2. If that option is not acceptable, the tie-down strap system could be used with a larger gap between the back of barrier base and face of raised sidewalk, say 6 to 9 in. and at TL-2. Finally, it might seem reasonable to use a combination of the tie-down strap system and the WisDOT tube/saddle system at an even closer offset to front face of sidewalk, say 3 to 6 in., for TL-2 impact conditions of NCHRP 350. Actually, I might consider the latter option more preferred when considering the 5-year work-zone time period. Of course, it should be noted that these options have not been tested under TL-2 conditions and/or when positioned this close to a raised sidewalk edge.


Finally, one last consideration would be to further reduce posted speed limits from 45 mph to 35/40 mph to help control potential impact speeds.


If you have any further questions regarding the enclosed information, please feel free to contact me at your earliest convenience. Thanks!

Steel Posts in the Bridge Transition Section

State: IA
Date: 08-06-2015

We received a call from a contractor asking if they could
use a longer 6'9"  for posts 8-13 in that attached modified road
standard.  This would keep the same length of embedment but would allow
the top of the post to be flush with the blockout.   It sounds like
they are having trouble getting the post installed correctly with the method
they typically use.  The other reason they sighted, for the longer posts,
is that the dies their manufacture have need 7 inches at the top of the post to
punch the holes.  The shorter posts shown are requiring custom punching of
the posts and driving costs up.  Do you see an issue with allowing the
post to be 3 inches longer?   They are waiting to order some posts,
so a quick response would be greatly appreciated. 

Attachment: https://mwrsf-qa.unl.edu/attachments/6668147e0cd2bef85398813a1ac55ad0.jpg

Date: 08-07-2015



I have reviewed our prior research studies that pertain to this issue. These include:


System No. 1 - ITNJ Test Series – nested thrie-beam AGT with steel posts on ¼-spacing near buttress end

Report can be accessed at:


Test ITNJ-2 on an improved design consisted of successful 2000P pickup truck test at TL-3 impact conditions. The six ¼-spaced posts were installed with a 49" embedment depth and a 2" recess for top of post below top of 31" tall thrie beam rail. The distance between the buttress end and the center of the first post was 11.5". The top of posts were recessed 2" in this stiffer transition region due to concerns for engine hood and quarter panel contact and snag that may lead to instabilities and even rollover. This snag behavior was observed and deemed a contributing factor in the rollover of a pickup truck in a R&D study performed on a Missouri transition system. See System No. 2 discussion below. This testing was performed under NCHRP Report No. 350.


System No. 2 – MTSS Test Series – dual 10-gauge thrie-beam median AGT with steel posts on ¼-spacing near buttress end

Report can be accessed at:


Test MTSS-2 on a modified design consisted of successful 2000P pickup truck test at TL-3 impact conditions. The six ¼-spaced posts were installed with a 43" embedment depth and a 2" recess for top of post below top of 31" tall thrie beam rail. The distance between the buttress end and the center of the first post was 11.5". Following the first failed test, the top of posts were recessed 2" in this stiffer transition region due to engine hood and/or quarter panel contact and snag on top of the median posts and on the upper sloped end of concrete buttress. This testing was performed under NCHRP Report No. 350.


System No. 3 – Test 2214T-1 – nested thrie-beam AGT with steel posts on ¼-spacing near buttress end

Report can be accessed at:


Test 2214T-1 consisted of same system evaluated under System No. 1 above and included a successful 2270P pickup truck test at TL-3 impact conditions. The six ¼-spaced posts were installed with a 49" embedment depth and a 2" recess for top of post below top of 31" tall thrie beam rail. The distance between the buttress end and the center of the first post was 11.5". The top of posts were recessed 2" in this stiffer transition region due to concerns for engine hood and quarter panel contact and snag that may lead to instabilities and even rollover. This testing was performed under MASH.


System No. 4 – nested thrie-beam AGT to T131RC using steel posts on ¼-post spacing [no recessed posts]

Report can be accessed at:


Two tests - 2270P and 1100C - were successfully run under MASH TL-3. The six ¼-spaced posts were 7-ft long but did not utilize a recessed top of posts relative to the top of the rail. As such, this testing may suggest that the recessed region may not be needed under MASH.



**At this time, I have requested two other crash test reports from TTI. I want to review these tests and determine whether those successful thrie beam AGT systems utilized recessed tops for posts relative to top of rail. I should have those reports by next week. In closing, I believe that there is strong potential for the tops of closely-spaced, steel transition posts to incorporate the same height as the blockouts under MASH testing. However, I want more evidence from a few more cases for confirmation, and then I will get back to you with an updated response.


Thanks again and feel free to contact me with any additional questions or comments!

Date: 08-27-2015

Hi Ron,


Sorry for the delay in getting you this information.  It is amazing how quickly things get buried these days.  I appreciate the reminder. J


The first test was run under NCHRP Report 350.  It did not have a curb, and the end shoe was rotated into the slope of the concrete bridge rail parapet to which it was attached.  The link to the report for this test is:




The second test was run under MASH.  It also did not have a curb.  However, in this case, a special adaptor block was fabricated for use under the end shoe to keep the thrie beam rail in a vertical plane throughout its length.  The link to the report for this test is:




Please let me know if you have any questions or need any additional information.   We can forward video, etc. as needed.


Best regards,



Date: 09-09-2015
I believe that the recess may not be needed based on System No. 4 noted below. Thus, an increase in post length without the use of top recess may seem reasonable.

I agree that you could move forward with the longer post that no longer includes a recessed top region.

NHSX-52-5(31)--3H-96 Curved Guardrail

State: IA
Date: 08-07-2015

As we discussed please see the
curved guardrail layout.  I have include the MicroStation files so you can
do more detailed measurements.   





As I eluded to
below and again on the phone, I think a radial guardrail design needs to be
used for both the Pulpit and Madison roadways. Looking at the attached Google
Earth image, one can see the (no longer) existing layout changes from steel to
wood posts around the curve. We need to effectively mimic that design this time


I contacted
WHKS to get their design file as the plan sheets leave a lot to be desired.
Attached is their reply. Those files and all workup files are available at W:HighwayDesignMethodsSectionMethods-SubjectAreasBarriers_DesignModificationsGuardrail_RadiusGuardail_US52


‘Guardrail Details' inside 7288.14.dgn shows that the proposed paved
shoulder at Pulpit has an inside radius of 46' (attached image). Using that as
the design radius (as it is already poured is effectively becomes the radius),
the curved portion of the guardrail layout is shown in attached CurvedDimensions.pdf.


With an Lg of
75, you'll need (75/6.25) + 3 = 13 CRT posts and (75/12.5) = 6 curved w-beam
pieces. You'll also need to add a 12.5' VT section between the end of the last
curved section and the beginning of the End Terminal. The 3 posts that are used
in that section are the 3 added in the sentence above.


Using model
‘Guardrail Detail Shading', the guardrail splits should look something like

BTS= 28.125

VF = 75

VT1 = 250

Curved VT = 75

VT2 = 12.5

ET = 50

I'm not
entirely confident in the precision of the first VT as the paved shoulder and
guardrail don't seem to run parallel to each other in the 7288.14.dgn, but
since the paved shoulder is already constructed, drill the holes at 6.25'
increments and begin placing CRT posts at the junction of the ~250' VT and the
curved VT, continuing through the curve and final VT, ending at the end
terminal. The posts in the BTS and the VT1 may be changed to steel as requested
as I don't believe that change impacts the radial component.


For the Madison
Road curve, since it is only a partial layout, I can't really design it. The
same approach would apply though. Basically, the CRT posts begin at the
connection of the VT and the curved w-beam and continue through the curve.
Since the guardrail doesn't terminate near the intersection, continue the CRT
posts for two posts beyond where the curved pieces flatten out. You'll have to count
up the number of CRT posts in the field.


US52_CurvedGuardrailAtSideRoad.pdf is a highly modified detail of past
installations. Most of the information has been stripped out as the tables
didn't cover this large of a radius. The CRT hole spacing is borrowed from the
current BA-211.


You mentioned
over the phone that because of the severe drop off, additional actions were
being added in an attempt to extend the 10:1 grading at least 3' behind the
face of guardrail, with the potential of 4' before it broke to 2:1 or steeper.
Normally we would introduce longer posts for that situation, but since these
are CRT posts, I don't think that makes sense to do so. Brian and Dan may have
a comment on this.


I'm out of the
office tomorrow and on the road Monday, so I'll be happy to check back in on
Tuesday to see what Brian, Dan, and yourself decided to do for these
installations, including any necessary changes to the curved and VT lengths.




Image 806 is where we plan to
begin the ET,  which is 802 plus 25' as you note.  The side road this
is at is Pulpit.  This is an area that rock drilling needs to occur, and
Lovewell is marking out the post locations today so yes we need to know shortly
if the post spacing changes.




I would much prefer starting the end terminal from the
end of the rail shown on the attached 806 than the attached 802. It doesn't
appear the end terminal even terminates within the shoulder if starting from
the 802 measuring tape. I would concur that adding the additional 25' from the
802 to end up at the 806 and then start the end terminal would be the preferred


The trailing back around the radius of the side road
throws me a bit here. Looking at page D.3
of the plans
, it appears that they are running normal guardrail around the
corner of Pulpit Rock Road, if that is the location of these images, but I
don't see a detail in the plans for laying out guardrail on that tight of a
radius. We typically use a short radius guardrail detail that has different
posts and spacing at that kind of a radius.


I'm sure you're waiting on an answer, so I'll talk with
Brian this afternoon. In the meantime, can you provide the side road these
pictures were taken at if it isn't Pulpit/Madison, both of which have
potentially the same issue.




The guardrail layout is a long section that extends from
a bridge endpoint on the trailing end back around a radius of a side
road.  The section shown in the attachments is within the VT-2 section
with an additional 25' added.  If we didn't add the 25' the ET section
would end up flaring toward the side road.  Is this acceptable?



These pictures show the last two sections extending the
VT-2 by 25'.The ET would then be in the tangent paved shoulder area. 

Attachment: https://mwrsf-qa.unl.edu/attachments/16a1a95a1f898971310814814569c13c.zip

Date: 08-07-2015

A few years ago, MwRSF conducted a research study for the Wisconsin DOT. This effort explored the performance of W-beam short-radius guardrail systems under TL-2 impact conditions with larger radii. See the link to access a copy of the report. Also, note the Chapters & Sections that are recommended for reading.


See page 199 or PDF page 213!

See Chapter 12 – page 205 – PDF page 219!

See Chapter 13 – page 223 – PDF page 237!

See Chapter 14 – page 227 – PDF page 241!


From this simulation effort, it was determined that rail heights greater than 27 in. and up to 31 in. would improve barrier performance for pickup truck impacts. Although no small car simulations were performed, there is concern that a 31 in. rail height could accentuate small car underride. As such, it was believed that a 29 in. rail height may still provide improved performance for pickup truck impacts but reduce concerns for small car underride. In the absence of an actual crash testing program at TL-2, MwRSF personnel would lean toward the use of a 29 in. rail height versus a 27 in. rail height based on the best available information and results from this study. Of course, the only true evaluation of safety performance would be through full-scale crash testing.


Second, the study noted that blockouts on posts around the radius contributed to improved vehicle capture by better maintaining adequate rail height. Blockouts also showed an ability to reduce vehicle to post contact. Further, CRTs were simulated around the nose through the tangent sections. As such, it would be recommended to maintain the CRTs throughout the entire curve and into tangent for any larger radius system that is implemented.


It should also be noted that the simulation effort was performed with level terrain behind the barrier system. Your real-world scenario will likely feature a gradual slope behind the barrier system for some distance, followed by a steeper slope. Barrier performance can be greatly affected by the presence of various slopes behind the actual barrier. Thus, it is recommended to provide a gentle slope behind the barrier using as much lateral distance as feasibly possible.


Again, these thoughts are provided based on our best available information as well as the research findings from the recent simulation effort. If you have any questions regarding this information, please feel free to contact either myself or my included colleagues at your convenience. Thanks!


Non-Proprietary Bullnose Thrie-Beam System

Date: 08-17-2015

I was in Virginia a week ago and saw the Bullnose Thrie Beam in
a gore.  They had several.


While I have seen several in medians, this was the first time I
saw one in a gore. 


Attached are some pictures from Google of the location. 


Are the Midwest states doing this design in gores. 


Any pointers for this application, concerns?


What are your thoughts? 


Date: 08-17-2015

Thanks for the information. I want to give some background. As you know, MwRSF developed and tested this system in the late 90s. At that time, we tested a narrow system with parallel sides and requested eligibility of two wider alternatives with parallel sides to accommodate different median/hazard widths. During the review process, Dick Powers suggested an alternative design that included non-parallel sides (i.e, the back-side rail flaring away from the front rail that runs parallel with the traveled way. As such, the FHWA letter included an alternative layout that included Dick's suggested variation that is similar to the front of the system contained in your photographs.


Again, the as-tested configuration utilized parallel sides. The as-tested design could be used to shield the hazards shown therein and would result in shallower impact angles along the sides of the bullnose. The currently-depicted dual flares would seem to potentially increase approach angles for 1 or both sides.


In such scenarios, I would always prefer to use the bullnose in a parallel configuration if site conditions allow. However, the Power's alternative may allow for a flared version to be used in these settings when traffic is on both sides. One potential item to consider is the effect of flare angle on sides and resulting increased I.S. Historically speaking, there has not been considerable crash testing performed on crashworthy systems that are now installed with the maximum allowable flare angle provided in the AASHTO RDG for highway and WZ applications. We performed some testing on a flared MGS many years ago. However, PCBs/TCBs and some other devices may not always have had the allowable flare built into the testing program on the front end. That topic may be something to reconsider moving forward under MASH.


I have also copied Bob on this reply as he was largely responsible for the original bullnose system. He has fielded the majority of the bullnose implementation questions and may be able to provide additional input into this special scenario. Please let me know if you have any questions regarding the information provided thus far. Thanks!

Date: 08-18-2015

Ron made most of the good points on the bullnose in gore areas, but I have a couple more.


As Ron noted, there are concerns with flaring the bullnose sides on a couple of levels. First, is the increased IS issue Ron noted below. Second, is that we don't want the flaring of the bullnose to negatively affect the capture and energy absorption of the system. Thus we have typically not recommended flaring of the bullnose prior to post no. 5 on each side. Dick Powers did approve a flared version as noted below, but that was intended for medians and not necessarily gore areas where the traffic on both sides of the bullnose is in the same direction.


Also because the system in this application would have traffic in the same direction on both sides, the guardrail splicing would be different than a median bullnose installation in that the overlap of the splice on the left side of the system would be reversed.


We have addressed this issue in the past through our Pooled Fund Consulting efforts and I have attached those responses as well for you to review as they contain some additional thoughts.






As Ron noted, if you have any questions, let us know.

grading behind a MGS transition

State: WI
Date: 08-21-2015

If we have the 2' of relative flat grading behind the post of the MGS thrie beam transition, How steep can the slope behind the 2' of flat grading?  I assume a 2:1 would be acceptable.

Date: 08-21-2015
A 2:1 slope located 2 ft behind the posts (on level grading) should not cause any adverse affects to the performance of the guardrail transition.

Vertical Taper Rate for Concrete Barrier

Date: 08-26-2015

I have been
searching through various standards and reports trying to find documentation on
acceptable vertical flare rates when transitioning from 32" high safety shape
barrier to 42" safety shape barrier, or adjacent to thrie beam structure
connections to 42" bridge rail . As you will appreciate, many different slope
ratios are used by various agencies ranging from 1.5H:1V to 10H:1V. While
reviewing TRP-03-300-14, I saw the photo below of the TCB to permanent concrete
barrier, which referenced TRP-03-208-10.

Based on
dimensions of the steel cap rail transition in TRP-03-208-10, the slope is
4.97:1 (1262mm run over 254mm rise), or say 5H:1V. Can you offer any comments
on the rationale for this ratio to minimize snagging potential for the vertical
transition from a 32" PCB (pinned) to 42" permanent barrier, and whether using
5H:1V transition for a MASH thrie beam connection to 42" bridge rail would be
acceptable (ie 32" at end of bridge rail where nested thrie beam overlaps the concrete,
and immediately transitioning the top of concrete upward at 5H:1V to 42" bridge

Attachment: https://mwrsf-qa.unl.edu/attachments/49c33a8d68c89e96f5eac8697b32ec40.png

Attachment: https://mwrsf-qa.unl.edu/attachments/71e2ebccb8564ba33ea8c5563e7613e3.png

Date: 08-26-2015

You are correct that in TRP-03-208-10 we utilized a 5:1 vertical taper to go from 32" to 42". The 5:1 height transition slope that we tested with the PCB transition used a steel cap to create the vertical transition flare. There is some concern that using concrete to create the 5:1 flare may increase friction and or gouging of vehicle components in the flared region. Thus, the 5:1 is an aggressive approach with some concerns for its use. However it has been adopted by many states. Use of an 8:1 slope has been commonly used as a more conservative approach.


Based on the performance of the this 5:1 flared cap, it would seem reasonable to use slopes as high as 5:1 when transitioning from 32" tall barriers up to higher heights. We could not recommend the use of these higher flares for shorter barrier heights below 32" as the potential for the vehicle to climb the flared section may increase if the starting height of the flare is lower.


For transition from heights lower than 32", the recommendation would be an 8:1 slope. We would recommend this based on the concerns noted above regarding the difference in the slope materials. In addition, we would not want to go to steeper slopes for barrier heights below 32" on the low side due to concerns for increased vehicle exposure to the slope and climb.

Date: 08-27-2015

I note in the package you just sent me that the proposed Standardized AGT Buttress uses all W6x8.5 posts (with last 6 posts immediately adjacent to the concrete rail being 78" long and spaced at 18-3/4"). We are currently using the simplified steel post MGS stiffness transition with three W6x15x84" posts immediately adjacent to concrete rail spaced at 37-1/2" with next four W6x9x72" at 14-3/4", etc (per MWTSP similar to Missouri Transition to Single slope in TRP-03-210-10).  Has the post configuration in the proposed Standardized AGT Buttress with all W6x8.5 posts been crash tested before? Were there concerns with the performance of the MWTSP design using the W6x15 posts, or is the proposed change to all W6x9 posts in the proposed AGT buttress being primarily driven by desire to only use one size of posts (albeit two different lengths)? 

Date: 08-28-2015

I have a some answers to your questions.


First , the upstream stiffness transition that was developed in TRP-03-210-10 was designed to work with a wide range of AGT designs that were tested on only the downstream end in NCHRP 350. When we conducted the crash tests of that upstream transition, we selected a very stiff AGT design adjacent to the bridge rail. The system that was selected was a Missouri AGT to a steel bridge rail with a cap rail because it was a very stiff design that would accentuate any issues with the upstream end of the transition.


That said, that Missouri AGT with W6x15 posts at 37.5" spacing was never tested and evaluated with a concrete parapet. However, you are connecting to a concrete parapet. Thus, you may want to consider designing the downstream end of the transition to comply with a previously tested and approved transition for the downstream end. We provided guidance for adapting existing transitions for use with the upstream end that we developed in the project report (TRP-03-210-10). Thus you could adapt one of those designs, or we could assist you in adapting something else that you prefer to use.


Another potential option would be to use the W6x15 posts at half post spacing that you originally had and connect to a concrete parapet. I went through some old correspondence that we had with Iowa. They requested guidance on using the W6x15 posts at half post spacing and connecting to a vertical concrete parapet. We had replied then that the system was likely to work, but that there were concerns with slightly more thrie beam deflection relative to the more rigid bridge system. However, we believed that the increased deflection would pose minimal risk for wheel snag, excessive barrier deflections, or vehicle pocketing. This design would also require flaring of the end of the parapet to prevent snag. Use of this design would likely require further investigation into the relative deflection and snag potential of similar systems to justify its use. We recommended to Iowa to run this by FHWA as well. The upcoming testing on the standardized parapet may shed light on that as well.




The transition being used for the standardized parapet is a previously NCHRP 350 tested AGT design (with a curb) that was done for Iowa for connection to concrete bridge rail. It was selected for evaluation of the standardized parapet because it is among the least stiff of the approved transitions and it has been shown to be sensitive to use without or without a curb. Thus, we believed it will provide a critical test of the standardized parapet such that we can use it with all previously tested AGT's. The Iowa transition was also tested to MASH with the curb and passed during NCHRP 22-14.





So the Iowa transition has been around for a while, and the use of the W6x9 posts was not originally intended for simplification of inventories. However, during the development of the upstream stiffness transition, we were asked to considered simplified post configurations to limit inventories, and the Iowa transition was part of that thought process. 

Minimal guardrail lengths to develop tension over culvert

State: IA
Date: 08-26-2015

We have a
situation where guardrail is being attached to a short bridge and is acting as
the bridge rail, though we also see similar issues with culverts, and there is
a question regarding how long of an installation is needed on either side of
the bridge or culvert to develop enough tension in the system. I can somewhat
gather that for a TL-3 system, 62.5' is the minimum needed on the upstream and
downstream ends beyond the obstacle, assuming a typical end terminal as shown
in the attached standards. Please confirm or provide guidance.


The project in
question is leaning towards a TL-2 installation (gravel road, very low volume)
and the same question remains; what would the minimum combined length of w-beam
and end terminal typically need to be to correctly anchor and tension the
system for a TL-2 installation?


For a very
generic layout, see attached.


Thanks for your


Attachment: https://mwrsf-qa.unl.edu/attachments/d0b01a764b637e917d16273edb2ea2f7.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/3747c7df94a35634eec4259539f29241.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/d2ed67ce4498074a7b14264d3d3d3d6a.pdf

Date: 08-26-2015

We have looked into minimum system length for the MGS in the past as well as made recommendation regarding minimum system lengths for long span guardrails over culverts with reduced span lengths from the 25 ft unsupported span that was crashed tested. There are three factors that come into play.


1.       Lateral Extent of the Area of Concern, the Guardrail Runout Length, and length-of-need (LON) as determined by the Roadside Design Guide (RDG) must be considered when factoring in minimum system lengths. Often a guardrail system may be able to redirect vehicles at lengths less that that required to adequately shield the hazard. As such, determination of length should start here. If the runout lengths are short enough to consider s shorter barrier system, then a couple of other factors need to be considered.


2.       Minimum system length required for capture and redirection.


The MGS Long-Span Guardrail System over culverts was successfully crash tested and evaluated according to the Test Level 3 (TL-3) safety performance criteria found in MASH. For this testing program, the overall system length was 175 ft, including 75 ft of tangent rail upstream from the long span, a 25-ft long unsupported length, and 75 ft of tangent rail downstream from the long span. As part of the final recommendations, MwRSF had noted to provide a minimum “tangent" guardrail length adjacent to the unsupported length of 62.5 ft. While your installation does not appear to use the long span system, similar logic may apply.


A recent MASH crash testing program on a minimum length version of the MGS suggests that there may reason to consider potentially reducing the 75-ft total guardrail length on the upstream and downstream ends of MGS Long-Span Guardrail System.


In the minimum length study, computer simulation and full-scale testing indicated that a 75' long MGS system would be capable of redirecting a 2270P vehicle under the MASH TL-3 impact conditions. Test no. MGSMIN-1, was performed on the 75-ft long MGS with a top rail mounting height of 31 in. A 4,956-lb pickup truck impacted the barrier system at a speed of 63.1 mph and at an angle of 24.9 degrees. The test results met all of the MASH safety requirements for test designation no. 3-11. The tested system had a total of 13 posts.


A performance comparison was conducted between 75-ft MGS (test no. MGSMIN-1) and 175-ft MGS. The dynamic deflection for the 175-ft (53.3-m) MGS was slightly higher than observed for the shortened system, but this difference could be due to variations in soil compaction between tests. The working width was nearly indistinguishable. In general, the 75-ft MGS in test no. MGSMIN-1 performed as desired and closely resembled the standard 175-ft MGS.


A second study regarding downstream anchoring of the MGS found that the MGS would successfully redirect 2270P vehicles impacting at 6 posts or more upstream of the end of the system for a MASH TL-3 impact on a 175-ft long MGS system. 


Based on previous testing and the results of test no. MGSMIN-1, MASH TL-3 vehicles impacting between post nos. 3 and 8 of the 75-ft long system should be redirected. Vehicles impacting downstream of post no. 8 may be redirected, but the system would also be expected to gate based on the downstream anchor research.


Based on the MASH 2270P test into the MGS Minimum Length System, we believe that the MGS Long-Span Guardrail System would likely have performed in an acceptable manner with 62.5 ft of rail on the upstream and downstream ends, thus resulting in an overall system length of 150 ft. A 62.5-ft long tangent length adjacent to the unsupported length would still provide adequate space to incorporate a 37.5 ft or 50 ft long energy-absorbing guardrail end terminal.


For unsupported lengths of 18.75 ft and 12.5 ft, it would seem reasonable to consider a reduction in the required guardrail length both upstream and downstream from the unsupported length using the test information and arguments noted above. For two missing posts or an unsupported length of 18.75 ft, we believe that the upstream and downstream guardrail lengths likely could be 56.25 ft each with a minimum overall system length of 131.25 ft. For one missing post or an unsupported length of 12.5 ft, we believe that the upstream and downstream guardrail lengths likely could be 50 ft each with a minimum overall system length of 112.5 ft. However, we believe that the three CRT posts still would be required on the upstream and downstream ends of the 18.75 ft and 12.5 ft long unsupported lengths. In addition, one would need to discuss with and likely obtain approval from the manufacturers as to whether they would allow three CRTs to be used within the last 12.5 ft of a 50-ft long guardrail terminal.


If one were to follow the logic used above and consider the situation of no missing posts (i.e., 6.25 ft post spacing throughout), the upstream and downstream ends would be reduced by 6.25 ft each and include the interior 6.25 ft long span in the middle of the system. As a result, the overall system length would be 43.25 ft + 6.25 ft + 43.25 ft for a total of 92.75 ft. As noted above, MwRSF recently crash tested a 75-ft long version of the MGS with satisfactory results, effectively configured with two 37.5-ft long guardrail segments with tensile anchorage devices and placed end-to-end. This corresponds to the situation in the schematic you sent and would provide conservative guidance on minimum length for the guardrail system over the culvert. Thus, this would correspond to 43.25 feet of barrier on the upstream and downstream end of the system. However, some terminals may require a 50 ft length for installation.


Of course, it should be noted that these design modifications are based on engineering judgment combined with the unpublished results from the MGS Minimum Length System crash testing program. In addition, the opinions noted above are based on the assumption that the currently-available proprietary guardrail end terminals would provide comparable tensile anchorage for the MGS as provided by the common tensile anchorage system using in the MwRSF crash testing program (i.e., two steel foundation tubes, one channel strut, one cable anchor with bearing plate, and BCT posts at positions 1 and 2 on each end). Although we are confident that the modifications noted above would provide acceptable performance, the only sure means to fully determine the safety performance of a barrier system is through the use of full-scale vehicle crash testing.


3.       Sufficient length for compression based terminal operation must be considered as well.


To the best of our knowledge, the shortest installation lengths for compression based terminal testing was conducted on 131.25-ft long system. We believe that this length could be shortened some based on our current knowledge of guardrail compression forces. We have used a reduction in longitudinal rail force of approximately 1-1.2 kips at each post in a guardrail due to the connection between the post and the rail. Current terminal designs tend to have impact head compressive forces that average about 15 kips. This would mean that a minimum of 12-13 posts would be needed to develop the compression load. Of course the end terminal takes out some posts during its compression. However, most of the velocity drop occurs in the first 25-31.25 feet of the compression. Thus, we can assume that if we allow for 31.25 ft of compression and 13 posts to develop the compressive load, an estimated minimum system length for the development of the end terminal compressive loads would be 112.5 ft (13*6.25+31.25).


Because we did not have additional funds or terminal testing and evaluation in the above research, we would recommend minimum system lengths of 112.5 ft in order to be conservative.


One last factor to consider with the use of terminals on these short systems is the deflection of the terminal when impacted on the end relative to the hazard. As noted above, we believe that the system will redirect the vehicle beginning at post no. 3 in the system. However, in an end on impact of the terminal, the vehicle may deflect down the rail between 37.5 ft – 50 ft. Thus, hazards near the back of the guardrail may still be impacted by end terminal impacts even when they are in the redirective area of the guardrail system. As such, you have to consider both the deflection of the terminal, the redirective region of the LON, and the runout length considerations when designing the placement of short guardrail system.


Thus, based on the analysis and review of previous research, it seems that the minimum length of the installation may be limited to 112.5 ft based the function of the compression terminals. In answer to your question with respect to guardrail over culvert, we would recommend that the overall system length be at least 112.5 ft, and that a minimum of 43.25 ft be required on the upstream and downstream ends of the system. Again, consideration of Lateral Extent of the Area of Concern, the Guardrail Runout Length, and length-of-need (LON)may trump this guidance.



travesable cuvert grates

State: WI
Date: 09-02-2015

We had some field staff have a problem installing a steel traversable culvert grate on a concrete pipe (see photo). 

We and a number of other states allow the use of an adapter to connect the steel traversable culvert grade to concrete pipe. 

I was wondering if MwRSF has seen similar problems in the past or has come up with other possible solutions (e.g. traversable concrete grate comes to my mind...).


I know that the 4" object standard is difficult to apply on a slope, but it is the best published  guidance I know of 

Attachment: https://mwrsf-qa.unl.edu/attachments/2a74b49be0867b63be4149f6a5867240.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/35392f576bbdd3346a8bb66f4b2f947e.pdf

Date: 09-14-2015
I have a couple of comments regarding the attached detail. 

1. I assume that the detail is for parallel drainage structures based on the pipe alignment which is why you are concerned with the step up at the attachment to the culvert pipe. If so, I would note that the detail shows 4:1 slopes for certain configurations. Currently, the best available guidance for traversable parallel drainage in the RDG recommends maximum slopes of 6:1 at speeds up to 80 km/h. 
2. With respect to the step up in height at the top of the culvert, the current best guidance for traversable parallel drainage would suggest that this is acceptable. Currently, the RDG recommends that the height of the first pipe on the culvert be mounted 4-8" above the culvert invert. Thus, if the slope grading remains consistent with the traversable culvert grate, then the step of of 6-8" at from the top pipe of the grate to the lip of the concrete pipe should present a similar traversable step. 

Let me know if you have further comments or questions.


Guardrail deflection for non-TL-3 situations

State: IA
Date: 09-14-2015

We've had the following issue come up recently and I'm wondering if you can provide some guidance as to what to use for design deflection distances.

The attached Table 5-6 Summary of Maximum Deflections from 2011 Roadside Design Guide provides design deflections for TL-3 impacts. What I'm wondering is what should we be using for deflection distances on a TL-2 situation, commonly an urban 35mph roadway? This question comes up when protecting a railroad lighting/crossbuck pole just off the traveled way, see attached. We currently use the attached BA-204 as the downstream anchor and the attached BA-253 as the general layout. Since the BA-204 utilizes a cable, introducing half post spacing is likely not preferred. That would leave us with nesting the thrie-beam if we felt it was needed.

Are there any reports that might give an indication if we should stay conservative and use the 19.2" number, as I think a 25 impact angle is highly unlikely, or use something more towards the 13.1" number? I'm trying to generate a balance between being far enough away from the railroad pole but not get too close to the roadway to introduce unnecessary impacts.

As always, I appreciate your assistance and insight.

Attachment: https://mwrsf-qa.unl.edu/attachments/c0ce28319a79f5948119e30f27f5321e.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/bd8b0eef399bf06ad91cfe111d87a5dc.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/4f4b2d63eddeb0ad9ae6aa2923b0bfe4.jpg

Attachment: https://mwrsf-qa.unl.edu/attachments/6a60e5ef20a3f7cc906617ff1110222c.pdf

Date: 09-14-2015

I have  looked at the details that you sent and have concerns over the function of the shielding of the pole outside of the allowable deflection.


As it is laid out in your detail, the signal pole is directly adjacent to the end anchorage for the thrie beam. Vehicle impacts in that area would likely release the end anchor through fracture of the BCT and allow the vehicle to impact the pole with little appreciable drop in velocity. Thus, shielding of the pole in this manner may not be effective.


That leaves a variety of options.

1.       Moving the pole upstream of the end of the thrie beam and away from the anchorage may help somewhat, but the longitudinal offset may not be allowable based on where the signal needs to be placed.

2.       Offsetting the pole more laterally may be the simplest answer, but I don't know what space restrictions you have for lateral offset. A pole with an L-shaped top that allows the pole offset to increase while maintaining signal position might be an option too, but I don't know if such an option exists.

3.       A more effective post shielding could be done using a short concrete parapet with a TL-2 AGT and end terminal. This may increase the system length, but it would alleviate concerns for interaction with the signal pole.

4.       Install a breakaway base on the pole to make the signal similar to a breakaway luminaire pole.


Take a look at these options and let me know what you think. I may have missed something.


MGS Thrie Beam AGT with Curb

State: OH
Date: 09-15-2015

I have a question about our MGS-3.1 Bridge Terminal Assembly, Type 1 that I hope you can answer. The standard drawing states that “Where curb must extend upstream of Post No. 11 for drainage purposes, an extra 12'6" panel of 12 gauge w-beam must be nested prior to the transition (upstream of Post No. 13). This added component shall be included as incidental to the cost of the BTA."


Is this still required if the curb is asphalt and not concrete? The Ohio Turnpike constructs concrete curb from structures to BTA post 11 and then changes to asphalt.

Date: 09-21-2015

The need for the nested rail section is based on additional loading of the rail due to the curb geometry affecting the vehicle as it impacts the barrier. Thus, the material that the curb is constructed of should not make a difference, and the nested section should still be required for an asphalt curb section.


Please feel free to contact me with additional questions or comments. 

Guardrail Connector Plate

State: MN
Date: 09-16-2015

MnDOT, is in the final phases of developing a new AGT standard. It's the thrie-beam version with the first three larger posts (sized at 84" – W6 X 15) before the concrete end parapet connection. See the attached proposed standard plan (694_AGT_type31_SingleSlope.pdf).
Most newer bridge designs will have an integral abutment with the approach panel. The expansion/contraction joint will be as the end of the approach panel and the parapet.
A concern has been brought to our attention concerning the thrie-beam anchorage plate (see attached standard 8350A). The concern is that the 1 1/8" slotted holes will not allow enough room for expansion/contraction and that the guardrail will be push back and forth, which would move the posts out of vertical. While we do not know if this is a valid concern, we do have an existing standard for a w-beam rail anchorage plate with 3" slots (see attached standard 8318C).
Our question; would a modification of our 8350A standard, to include a 3" (or less) slot, be acceptable for our proposed 694_AGT_type31_SingleSlope.pdf design?
I look forward to your response. Please feel free to give me a call if you have any questions or need additional information.

Attachment: https://mwrsf-qa.unl.edu/attachments/00f9e32d43bc9b3555499917fc3885e5.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/575de1d94ad6db47b4300e40ddf86c7c.pdf

Date: 09-17-2015

I have accumulated some feedback from my colleagues. Here are a few thoughts.


Karla Summary:

In the hardware guide, RWE02a-b is the W-beam Terminal Connector. When comparing the RWE02a-b slots to the slots in MNDOT's standard 8318C, they are similar. The hole diameter in the MNDOT drawing is a little smaller. Slot dimensions for RWE02a-b is 31/32" x 3", while slot dimension for MN 8318C is 29/32" x 3".


RTE01b is the Thrie-Beam Terminal Connector. When comparing the RTE01b slots to the slots in MNDOT's standard 8350A, they are different. The RTE01b shows the slots at a 50-degree angle, and they are 31/32" x 1¾".  However, it is noted that the slots could be oriented parallel to the longitudinal axes rather than at the 50-degrees. MNDOT's standard has a smaller diameter and a shorter length, 13/16" x 1â...›".


Therefore, I would say MNDOT could at least increase the slot size to what the hardware guide shows without any concerns. I know that this doesn't get to 3".


Scott Summary:

First thoughts – a 3" slot is large. During impacts, it may take quite a while for the rail to shift enough, develop tensile loads, and create the membrane action typically associated with guardrail redirections. This shift could lead to pocketing/snag issues. On the other side of the argument, transitions rely more on lateral force loads provided by posts and rail bending. So maybe enlarged slots don't cause any issues. Without tests, I'm unsure what effect this has on performance.


An idea to strengthen their transition with 3" slots may include shortening the unsupported length of the thrie beam between the concrete parapet and Post 1 (shift the guardrail system DS). Reduce this length from 29" to say 15"-18". I would feel better about reduced snag concerns on the concrete end with a reduced distance.


I would be more comfortable if the expansion/contraction joint had concrete on both sides of it – which I believe is more common amongst our Pooled Fund States. Ideally, the 7-ft long standardized buttress would be on the upstream side of the joint. However, they may be trying to shorten the length of the transition system and thus, have spanned the thrie beam across the joint. Not sure they want to add an additional 7 ft.


Last thought, I hope that any potential enlarged slots would be punched into the end shoe during fabrication, not cut by hand just prior to installation. Are they working with a manufacturer to produce the part?




Based on the above feedback, there are some concerns with extending slots more than used in historical crash-tested systems. Increased slot length could lead to increased lateral deflection in advance of the rigid end. With a longer slot at the end shoe splice location, the rail tension will not develop as quickly. I concur with Scott on this issue. Instead, the rail may more easily move back even though resisted by posts and rail bending capacity due to overlap on parapet. At this time, we unfortunately do not know how much longitudinal shift is acceptable before rail tension is developed. I am aware from prior BARRIER VII modeling efforts that rail tension can be greater than some expect and near the buttress end, say up to 80 to 90 kips during high-energy impact events. Excess rail deflections can lead to pocketing and/or wheel snag as well as increased propensity for vehicle instabilities.


Unfortunately, I cannot easily determine the exact slot lengths, and bolt positions within those slots, that were used in prior crash test efforts of AGTs. What we can say is that standard thrie beam end shoes were used with the commercially-available, industry-accepted slot patterns and sizes, including some variations denoted in hardware guide.


Finally, if I had a choice, I would rather locate the expansion gap between the concrete rail end and the buttress end. However, it may be possible that the 3-in. slot length will not cause any problems in a crash testing program. Unfortunately, I am less certain regarding the system's safety performance with a change under MASH.


Please let me know if you want to further discuss this issue! Also, I need to look at a few other dimensions in the near future. Thanks!

Date: 09-23-2015

We appreciate your (Karla and Scott, also) information and insight.


We are looking into using the TF-13 design RTE01b, and have a few questions.


Was the RTE01b version used in your research outlined in the Transition Report TRP-03-180-06?   We have this as one of our standard options, but could not tell from the report.


I have seen both versions on other state standards, one similar to ours and the RTE01b.


Also, I am interested in the number of attachment holes for attachment to the barriers on the end.  Ours has five 1" diameter holes and RTE01b has 9 (two ¾" and seven 1").

Why is there a difference.  When would we ever need more than the five 1" holes?


You help is appreciated.

Date: 10-01-2015

Hello Mike!


I will respond to your questions below. I have also provided a photograph from the noted reported to inform you of what bolt hole pattern was used in this test series. The end show have 5 bolt holes versus nine.


Photo from test 2, page 71, figure 46 of TRP-03-180-06.


Ronald K. Faller, Ph.D., P.E.

Director and Research Associate Professor


Midwest Roadside Safety Facility (MwRSF)

Nebraska Transportation Center

University of Nebraska-Lincoln

130 Whittier Research Center

2200 Vine Street

Lincoln, Nebraska  68583-0853


(402) 472-6864 (phone)

(402) 472-2022 (fax)



From: Elle, Michael (DOT) [mailto:michael.elle@state.mn.us]
Sent: Wednesday, September 23, 2015 11:39 AM
To: rfaller1@unl.edu
Cc: Brown, Timothy (DOT) <timothy.j.brown@state.mn.us>
Subject: RE: Guardrail Connector Plate




We appreciate your (Karla and Scott, also) information and insight.


We are looking into using the TF-13 design RTE01b, and have a few questions.


Was the RTE01b version used in your research outlined in the Transition Report TRP-03-180-06?   We have this as one of our standard options, but could not tell from the report.

**As shown above, the as-tested end shoe part used 5 mounting holes. It does appear as though the CAD details in figure 45, page 70, shows extra holes in the end shoe.


I have seen both versions on other state standards, one similar to ours and the RTE01b.

**The previously-provided TTI report revealed static testing on multiple versions of end shoes.


Also, I am interested in the number of attachment holes for attachment to the barriers on the end.  Ours has five 1" diameter holes and RTE01b has 9 (two ¾" and seven 1").

Why is there a difference.  When would we ever need more than the five 1" holes?


**Thrie beam end shoes are anchored with 5 bolts – three in column 1 and two in column 2 located 8 in. away from vertical centerline of column 1. Unfortunately, I do not know the history as to why some end shoes have 9 holes. I would suspect that a greater number of holes spaced close to one another could potentially lead to part fracture at slightly lower loads. At any rate, we use only 5 anchor bolts for these parts.


Let me know if you have any further questions regarding this information.

Bridge Approach Section

State: NE
Date: 09-17-2015

I have questions about the use of 29" bridge rail & 28“Bridge Approach Section (BAS) taper to 31" over 50'

Has any thrie-beam BAS been tested at 28"?
Should it work properly?

Date: 09-18-2015

We have been asked previously about reduced height transitions. We would expect that a slightly lower transition height would allow for redirection of the vehicle. However, all thrie beam transition testing that I have seen was conducted at 31" or at the metric height of 31 5/8". In addition, lower transition heights typically would expose more of the concrete barrier at the downstream end of the transition and create a potential snag hazard. This would not be an issue for you as you are using lower bridge heights as well.


With respect to transitions, the issue has not been adequately addressed with respect to the current MASH vehicles and impact conditions.


We cannot say with certainty that reducing the rail height of the MGS transition will not affect its performance. Reducing the height of the transition 2"-3" will increase the impact load height on the post and the relation of the vehicle front to the barrier. This in turn could increase the moment on the posts and affect the lateral deflection and stiffness of the system. In addition, there would be concerns for vehicle stability as well as an increased potential for wood posts in the transition to fracture. Thus, we would generally recommend a height of 31" for the MGS transition as well.


We have noted that for TL-2 applications, there is increased potential that a reduction in the rail height for the transition of two inches would likely be acceptable.


Let me know if you need anything else.

MGS Stiffening

State: WV
Date: 09-18-2015

Do you know of any research and results of stiffening MGS with double nested beams or reduced post spacing.

Of course with added post it becomes a splice on the post system, but I was interested in the transition from MGS to the changed section. I was looking at the RDG, Table 5-6 on Page 5-34 and the maximum deflection of a Double W-Beam and 38 in post spacing looks desirable.

I am protecting High Mast lighting in the median that has a 4' dia. foundation 2' above grade and a poles that will not break. I really want some solid protection and bullnose can't be graded in due to the bifurcated typical. 

Date: 09-22-2015

We have looked at stiffening methods for  the MGS system in the past. We have not evaluated nested rail applied to the standard MGS system at this time. We have used nesting in a couple of special applications for transitions, but we have not done it for a standard length of need system.


That said, we have looked at the use of reduced post spacing for the MGS. The report for this work can be found at the link below. In the report, we tested ¼ post spacing and developed guidance for ½ post spacing deflections as well.




Historically, common W-beam guardrail systems have been easily transitioned between full and half-post spacing variations as well as half- and quarter-post spacing configurations without changes to post lengths or rail configurations. When the MGS with quarter-post spacing is deemed necessary to shield hazards closer to the traveled way, the needs exists to connect full-post spacing MGS to quarter-post spacing MGS. Under this scenario, MwRSF has previously suggested that an intermediate stiffness transition be utilized to more gradually blend the varied lateral stiffness of the two systems. More specifically, MwRSF suggested that a 12-ft 6-in. long MGS segment with half-post spacing be used to gradually transition the lateral barrier stiffness and strength, thus resulting in four spans of half-post spacing between the two systems.


Although the standard MGS utilized mid-span locations for rail splices, it would be expected that rail splices would occur at post locations for the MGS variations which utilized a reduced post spacing. Thus, MwRSF has suggested that rail splices be configured to occur a minimum of 1 reduced span (3 ft - 1½-in.), and preferably 2 reduced spans (6 ft " 3in.), beyond the last or first MGS full-post spacing.


The stiffness transition noted above is suggested for situations where impacting vehicles first contact the full-post spacing MGS and subsequently engage the quarter-post spacing MGS. Therefore, it would not be necessary to apply a similar stiffness transition to the downstream ends of quarter-post spacing MGS unless prone to reverse-direction impacts.


As a side note, we currently have a research project with the Illinois Tollway to evaluate the minimum offset for luminaire poles behind the standard MGS system. We can keep you up to date on the outcomes of that study if you are interested.


Lateral Placement Tolerance for Guardrail Posts

State: OH
Date: 09-22-2015

I'm not sure how familiar you are with our old Type 5 guardrail, but I have a question regarding its installation. Over a long run of guardrail, we have one location where, because of an obstruction, the post spacing was changed in the field to 6' on one side and 6'6" on the other versus the standard 6'3" and holes punched into the rail to secure the rail to the post. Is there a tolerance for lateral placement of guardrail posts and if so, what might that be?
Attachment: https://mwrsf-qa.unl.edu/attachments/edf447ad82342663d5c9e53393628e2a.png

Date: 09-23-2015

For the older G4(1S) or G4(2W) systems installed at the original metric height (27.75") we don't believe that a 6" offset of a post should make a substantial effect on barrier performance. However, at some point, the alternation of the post spacing may become more of an issue due to the potential for pocketing, vehicle snag on the posts, and vehicle instability.


We believe that a 1 ft offset tolerance on guardrail post longitudinal placement is acceptable in discrete locations. We would not recommend throughout a system. Typically, we would expect the posts to be placed such that the tolerance provided by the rail slots would be sufficient for attachment of the guardrail. However, a single post offset 1 ft or less should not pose an issue in our opinion.


It should be noted that some care should be taken when field cutting a new post bolt hole such that stress concentrations that may reduce the rail capacity do not arise. Thus, the post bolt hole should be cut smoothly and correspond with typical post bolt hole dimensions. Additionally, we would recommend spray galvanizing the hole to prevent corrosion.


For the MGS system, we have successfully tests a single omitted post at standard post spacing under the MASH TL-3 impact conditions. Thus, for the MGS system, the tolerances and may be larger and/or the post may be simply omitted in that area. The report on that research should be out shortly.


Expansion and contraction of the guardrail connection to a bridge parapet

State: MN
Date: 09-24-2015

Two questions;

1. How concerned should we be with expansion and contraction of the guardrail connection to a bridge parapet.
As you know from our recent questions, this has become a hot discussion issue, but I would think that each post connection must allow some movement with the bolt and slot design, otherwise we would be seeing problems on long runs.

2. If we do move the parapet/end post off the integral approach panel abutment, then how much of a gap can we allow before we have to start considering a cover plate. Attached is a consultant concept that illustrates the concept. Note that I told the consultant that they would not be able to use this concept without an acceptance letter from the FHWA.

Attachment: https://mwrsf-qa.unl.edu/attachments/98fa5fe4ed0dc09322d28201dc2d80d8.pdf

Date: 10-01-2015

I will try to respond to your questions noted below.


First, I will begin with question No. 2. We re-examined the issue of a critical gap size or length. When considering 25-degree approach angles for passenger vehicles, we believe that excessive gap length could lead to increased vehicle snag at open joints of rigid parapets. Historically, we have used a 2 in. limit for allowable wheel overlap on the upstream end of rigid buttresses when associated with acceptable snag under thrie beam attached to approach guardrail transitions. The affiliated gap length would be around 4.3 in. Thus, I might suggest holding the gap length to 4 in. maximum.


I recall that this issue was raised many years ago following our testing of an open concrete bridge railing for the State of Nebraska. I looked through our old Pooled Fund Consulting Q&A site and found the following inquiry and response. Years ago, we also recommended a maximum gap length of 4 in. and the use of chamfered corners/edges.




With regard to question No. 1, we are somewhat concerned with large longitudinal movements in the steel guardrail near the bridge end. If bolts are located at the ends of rail slots, would it be possible for the rail to pull over posts in the same direction of weak-axis bending. Of more concern, we believe that excessive slot length at the buttress and end shoe location could allow for increased vehicle pocketing/snag on the buttress end as well as greater risk for vehicle instabilities during close impacts near the bridge but in AGT.


Please let me know if you have any further questions regarding the information noted above.


KDOT Thrie-Beam Standard Drawing

State: KS
Date: 09-24-2015

I was doing some research on the pooled fund site on a separate topic and came across a detail showing a wood post alternative for the MGS thrie-beam (see attached .png file). I thought this was still in development/hadn't been finalized yet. Can you confirm the “recommended wood post design for Tested MWTSP approach transition" details shown in the attachment would apply to KDOT's current MGS thrie-beam (standard drawing RD613A, also attached)? If we can have this as a wood post option can you confirm the blockouts for the thrie-beam are still 6" wide for the 8" wide x 10" deep 7'-0" long posts? Depending on your response I may put together some draft standard drawings for MwRSF review before KDOT implements them.
Attachment: https://mwrsf-qa.unl.edu/attachments/f01237c4703077f67a9c2c4d25be773e.png

Attachment: https://mwrsf-qa.unl.edu/attachments/6d6ae8f470f23cfc79cbd6def13bfe96.pdf

Date: 09-25-2015

A few comments:


The drawing for “Recommended Wood Post Design for Tested MWTSP Approach Transition"  (bottom sketch on the drawing page you forwarded) appears to be very similar to the steel post transition from RD613A.  You would just need to swap the 3 larger steel posts on the downstream end for the larger 8"x10"x6.5-ft wood posts shown in the drawing.  The blockouts can remain the same size – 6"x8"x19" (your email had a 10" height, but I assume that was a typo).


Just a comment/question on your current steel post transition…  What size posts are you utilizing for the three 7-ft long posts adjacent to the buttress?  I don't see a section note for the posts on the drawing set (other than on section C for the w-to-thrie transition element).  Nearly all transitions with 37.5"  post spacing adjacent to the rigid buttress require increased post sections, e.g., W6x15's, W6x25's, etc.).   I am unaware of any transition being designed and tested utilizing only W6x9 posts with a  37.5" post spacing.

Date: 09-26-2015
Thanks for the quick reply. I'm going to use TRP-03-243-11 as a reference to put the details together for a wood post option. KDOT's 7'-0" posts are W6 x 15 posts. That information is shown on a post details sheet, which I didn't send over yesterday. Once I have the draft drawings put together can I send them onto you for review?

Curb (4") in front of Concrete Barrier

State: MN
Date: 09-28-2015

We would like your input regarding on 4"
high curbs, particularly with respect to potential vehicle redirection.  There
is general language in AASHTO Green Book, the Roadside Safety Guide and the
Safety Manual, along with the MnDOT RDM, but nothing that I would consider


Many Trunk Highway and Interstate projects have included
these raised curb sections adjacent to shoulders, in order to manage snow
storage.  Typically, these configurations will have a concrete barrier
located behind the minimally raised curb section.  It has been my understanding
that a 4" curb has not been considered to be vertical curb ( with the
definition provided in the AASHTO Greenbook ) and the MnDOT Geometrics and/or
Safety Office has not identified these as a significant re-directional hazard
to vehicles.


I would appreciate your insight.


Thank You.


Minnesota DOT

Attachment: https://mwrsf-qa.unl.edu/attachments/4f0e8940a317b0076a3b59efc8bd2217.pdf

Date: 09-30-2015
With respect to the 4" curb alone, there are no issues with traversability or vehicle redirection. Various tests of curbs have been performed that have shown curb heights of 4" are not sufficient to redirect a vehicle nor do they tend to cause vehicle instability on their own. 

That said, the use of the 4" curb may affect the performance of concrete barrier designs depending on the offset of the curb and geometry of the concrete barrier as curbs adjacent to the barrier may increase vehicle climb and the potential for vehicle instability. 


State: KS
Date: 09-29-2015

Bob – I called and left a voicemail earlier today in reference to the following and attached: KDOT has had several discussions on this topic with MwRSF previously and it recently occurred to me there may have been a miscommunication. Previously KDOT has asked whether there is an MGS installation for this type of short radius guardrail installation, which there is not. The MGS terminology is a little misleading. I think what we really wanted to know was; is there a 31" tall version of this type of short radius guardrail installation? I think NDOR has details for this, but I wasn't able to find them on their website. I did some searching on the Pooled Fund Site, but didn't find an answer to my specific questions. I know you and Cody Stolle did some investigating on this topic through a Wisconsin funded study in 2014 and found the short radius installation performed acceptably, or maybe even better in some regards, at a height of 31" compared with the 28" height.

As a result I wanted to revisit this topic again. I've attached several draft standard drawings. RD619 is the current short radius guardrail installation KDOT uses for a mounting height of 28". RD619A is a modified version of that drawing keeping the hardware through the curved section the same, but transitioning to MGS on either side of the curved portion. To avoid confusion I'm referring to RD619A as a modified short radius guardrail installation mounted at 31". This is not an MGS installation, but transitions to MGS hardware on either side of the curved section.

My goal is to develop something that will allow KDOT to avoid height transitions within an installation and minimize the different types of hardware we are using in our guardrail installations while still maintaining performance comparable to other TL-2 configurations for similar systems. Currently our practice is to use the old 4G1S AGT with the old 4G1S posts and hardware all mounted at 28" (shown on RD619) when a short radius installation cannot be avoided. As a result we have locations where 3 quadrants of the bridge is all MGS hardware and one quadrant that is 4G1S, which KDOT refers to as the Conventional Guardrail System (CGS). The other approach we've taken where we have space is to install the MGS and then transition to a lower height of 28" over 50 feet up to the curved section of the short radius installation. At that point the hardware switches back to the short radius hardware for the remainder of the installation unless there is room on the end terminal side (along the minor roadway) to transition the height and hardware back to MGS. For additional information I attached the updated post details and the MGS AGT details, which I reference on RD619A.

Just to cover what I found on the Pooled Fund Site during my search I included a short summary below:

1. Response to IA dated June 26, 2012 not recommending the short radius mounting height be raised to 31". A height transition over 50'-0" from 31" to 28" was suggested. The issue we've run into at KDOT is these systems are used only when an intersection is in close proximity to a bridge or a needed guardrail installation. In the case of the bridges there often is not enough room (less than 50'-0") to transition the height so you end up with one quadrant using the old hardware.
2. I found a similar response to NDOR dated January 2, 2012 regarding the 31" mounting height.
3. I also read through TRP-03-296-14 Extending TL-2 Short-Radius Guardrail to Larger Radii. That report seemed to suggest, if I understood it correctly, mounting the short radius at 31" was acceptable for the configurations shown in the report.

The details I've attached are yet another variation on the details shown in the report I referenced in number 3 above. I'd like to discuss this over the phone at your earliest convenience before any thorough review is completed. We coordinated the height transition option we currently use when there is space available with MwRSF previously. The new attached proposed drawings would be identical to that type of installation with the height transition omitted since it would not be needed with everything mounted at 31".

Attached is a detail I was able to track down that NDOR is using. 
Attachment: https://mwrsf-qa.unl.edu/attachments/d613681324be40e5cea7cd9191e909b0.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/45a91280348fe5d285948504b9e5a219.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/339eae51de9991bb0bd500d732d34bcb.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/792b63264dac3443e1b558abdcdfd27d.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/370a3f204f2ccd08c7d3ae40e30c8951.pdf

Date: 09-30-2015

We currently are limited in what recommendations we can provide regarding short-radius type barrier systems. As you are aware, no short-radius system has met the crash testing criteria for MASH or NCHRP 350 at this time. TTI has recently done research on a MASH TL-3 thrie beam short-radius system, but to my knowledge that has not been approved by FHWA and we have some concerns regarding the test matrix and impact points used to evaluate that system. Thus, we are left with trying to make the best of the situation at hand.


The only short-radius system that has met FHWA eligibility is the 27" high TL-2 version of the Yuma county short-radius system that was analyzed by TTI. This is the system that has been implemented most recently by several states. Your system seems to vary from this design somewhat as it included additional cable anchorages near the nose section. I am not sure of the function of these additional anchorages, but you may want to reconsider them as they may not be consistent with any currently approved design.


Subsequent to that research at TTI, we conducted a simulation analysis for WisDOT regarding the performance of short-radius guardrail for larger radii under NCHRP Report 350. This study started with the Yuma county design and extended it to larger radii (over 25') This study found that the performance of 27" high short-radius guardrail was potentially limited in terms of capturing the 2000P vehicle and that 31" high larger radii short-radius systems had improved potential for pickup truck and high CG vehicle capture. The report also noted that the simulation analysis did not investigate small car interaction with the large radii short-radius systems at either 27" or 31" mounting heights. Passenger cars may underride the rail if a 31-in. mounting height is used despite a beneficial interaction with pickup truck vehicles. Previous thrie beam short-radius systems with 31-in. mounting heights culminated in small car underride and roof or windshield crush. No W-beam short-radius system has been tested and approved with a top mounting height higher than 27 in. Nonetheless, tangent guardrails as tall as 36 in. have redirected small cars at MASH TL-3 impact conditions. Based on these concerns, full-scale testing was highly recommended if a 31-in. (787-mm) tall system is to be used. Thus, while 31" rail height was shown to work acceptably in the study for a limited range of speeds and impact conditions, concerns were noted for small car capture that prevented us from fully endorsing a shift to 31". We did note that a 29" system may be a compromise between the two alternatives until further research is available.


As you noted in your email, we have made similar responses to Iowa and Nebraska regarding the height issue and have recommended limiting the height to the approved 27" for now even though we have some data that suggests that it may pose problems with high CG vehicle capture. This is due to concerns that the small vehicle capture may suffer. Thus we are currently limited to that guidance until further investigation or crash testing of the increased height short-radius systems are undertaken.


From your email, it appears that this is an issue because you are converting to the MGS system and the 27" height of the Yuma county system likely conflicts with some of the approach transition hardware for the MGS. Unfortunately, for the time being, we can only recommend the TTI/Yuma county system as it was granted eligibility at this time because we do not have sufficient information raise the height based on the concerns noted above. Thus, one may be forced to implement the current Yuma system and keep use older hardware.


Let me know if answers your question. I understand that this may not help much. The short-radius issue has been a big problem for several years and will continue to be until we can resolve it. We currently have and R&D effort with NDOR to evaluate a different treatment for intersecting roadways, but that work is still in the developmental phases. 

Attachment: https://mwrsf-qa.unl.edu/attachments/b2308bea4ad6c813fde235052ea0dde1.pdf

Date: 09-30-2015

I will revisit our existing design and compare it to the TTI design you provided in the attachment. Is there anyone at MwRSF who might be able to do a quick review of the details once I have them put together just to get another set of eyes on them in case I missed anything?

Date: 10-01-2015

Attached are the revised details. I removed all the details for the cable anchor assemblies, soil plates, etc. Since the TTI report noted any tested AGT/End Terminals meeting NCHRP 350 TL-2 or higher could be used outside the curved section I left the 4G1S AGT/End Terminals in the drawing. I attached the original drawing for comparison (RD619_Original). I also attached the other standard drawings I reference on the sheet for information. I did have a couple of questions:


1.    I didn't see this in the TTI report, but I know there was some discussion of this in the work you and Cody did for Wisconsin. Can the STYP posts be used in lieu of the wood CRP posts?

2.    I left the mounting height detail showing 28" rather than 27" since that is how tall KDOT's typical 4G1S w-beam is mounted. Is that acceptable?

3.    Are the radii listed shown on KDOT's version of the drawing ok? I know from the TTI report it had a radius of 8'-0" (shown as 7.96' on KDOT's version to give a length of 12'-6" for the rail, which is evenly divisible by the post spacing). From the work you and Cody did it looks like radii of 23'-10.5", 47'-9", and 71'-7.5" is also ok. Are radii increments between these radii acceptable? Can those radii be adjusted slightly to give lengths of w-beam rail that are evenly divisible by the typical post spacing (i.e. 23'-10.5" would be ok shown as 25'-0")?


Thanks for your help on this.


Attachment: https://mwrsf-qa.unl.edu/attachments/240e25b1dad7a78be175f1f569a629a4.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/fa1c06ec75ed9f118885f4df7c96edc0.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/9c6836d43c5f02d44ac420444796724b.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/414a5f15d82b26d488930407414fe5b9.pdf

Attachment: https://mwrsf-qa.unl.edu/attachments/0cf111e0c0400832e80830d739b72e7b.pdf

Date: 10-02-2015

I had another question that crossed my mind this morning after sending this info to you yesterday. Can the MGS guardrail and AGT be used instead of the 4G1S system outside the short radius installation if it is mounted at 27" or 28"?

Date: 10-03-2015

I have commented to the questions below in red.


Comments are listed here with respect to the details you sent.

1.       You details show a 2:1 slope starting 3' behind the short radius system. While I understand the reality of these slopes, no short-radius system has been successfully tested with these types of slopes and the approval of this system was based on level terrain testing. Based on previous testing and simulation done at MwRSF, we believe that the performance of the system will be degraded significantly with the presence of the steep slope behind the system in terms of vehicle capture and stability.

2.       I saw no other significant deviations from the TL-2 approved Yuma County system.



I had another question that crossed my mind this morning after sending this info to you yesterday. Can the MGS guardrail and AGT be used instead of the 4G1S system outside the short radius installation if it is mounted at 27" or 28"?


I don't see any reason why the MGS cannot be used mounted at the lower height for this application. The midspan splices should improve the performance and the deeper blockouts should aid in vehicle capture as well. Obviously the benefits of the reduced post embedment would not be included as the height would not increase. Cody's work for WisDOT indicated that the blockouts may improve vehicle capture for the higher CG vehicles. The effect of the blockouts on small car capture is unknown, but TTI noted in the Yuma TL-2 system report that blockouts could be used even in the curved section. However they did not make a recommendation towards the larger blockout in the MGS.  


With respect to the AGT, I think you would need to stay with the NCHRP TL-3 approved transitions noted in the TTI report. The MASH tested MGS transition essentially uses NCHRP 350 approved transitions on the downstream end adjacent to the bridge, so that part of the transition would not be different. The upstream end of the transition was designed to convert between the stiffness between the standard MGS and the AGT. It also used a asymmetric W-thrie transition piece. The upstream end of that transition has not been evaluated at the lower height, and if you recall, we experienced a rail rupture of that transition near the asymmetric W-thrie transition piece when we tested it with a curb and a 1100C vehicle, which forced us to nest the w-beam ahead of the asymmetric W-thrie transition piece when the AGT was used with a curb. Thus, a lower system may be sensitive to near the asymmetric W-thrie transition piece. Additionally, the asymmetric W-thrie transition piece would not allow for the correct thrie beam height for the AGT connection at the bridge.




Attached are the revised details. I removed all the details for the cable anchor assemblies, soil plates, etc. Since the TTI report noted any tested AGT/End Terminals meeting NCHRP 350 TL-2 or higher could be used outside the curved section I left the 4G1S AGT/End Terminals in the drawing. I attached the original drawing for comparison (RD619_Original). I also attached the other standard drawings I reference on the sheet for information. I did have a couple of questions:


1.    I didn't see this in the TTI report, but I know there was some discussion of this in the work you and Cody did for Wisconsin. Can the STYP posts be used in lieu of the wood CRP posts?

I am assuming that you are referring to replacement of the timber CRT posts used in the approved system with Steel Yielding Posts (SYP) developed by TTI. We did not comment on this in the WisDOT study, but we do not recommend replacement of the timber CRT posts with any of the steel breakaway posts at this time. The SYP post is a yielding post that bends at a lower load that the standard W6x8.5 section rather than breaking away at the base like a CRT. This behavior may create a ramp for the vehicle to climb in the nose section which could increase the propensity for override of the rail and vehicle instability. The UBSP post that was developed through the Midwest Pooled Fund is likely a better option as it was used successfully in the bullnose and tends to break way at the base. Component testing of that post compared well with CRT's. However, we have not recommended the use of that section in any system without full-scale testing as it may be sensitive to applications outside of the bullnose. In the case of the Yuma county short-radius system, it is unlikely that it will ever be subjected to a full-scale crash test to evaluate that potential application.

2.    I left the mounting height detail showing 28" rather than 27" since that is how tall KDOT's typical 4G1S w-beam is mounted. Is that acceptable?

I will leave the mounting height decision up to you and KDOT, because the guidance in this area is mixed. TTI received approval on the Yuma county system based on the 27" height. Thus, from the standpoint of FHWA eligibility, the 27" height has been recommended by both TTI and FHWA. As I noted in the previous email, the simulation effort we did for WisDOT showed found mixed results for the varying rail heights. Cody's simulations of a 5,000 lb pickup truck on the 27" high Yuma county system with an 8' radius that were conducted to validate the modeling effort found that the pickup was captured. However, a simulation of the a 4,409 lb pickup truck under the same impact conditions overrode the rail. Additionally, as we simulated larger radii at the 27" height, the 2000P vehicle vaulted over the guardrail in 100, 100, and 80 percent of impact conditions simulated for 24, 48, and 72 ft radii, respectively. Blockouts were added to the CRT posts, and the vaulting override rates were reduced to 80, 36, and 50 percent of simulated impact conditions for 24, 48, and 72 ft radii, respectively. Thus, while the 27" height was listed in the TTI report, our study found that it potentially may have capture issues with the higher CG vehicles. We also simulated 29" and 30" rail heights and found much better capture of the 2000P vehicle. However, increasing the rail height leads to concerns for small car underride which were not investigated in the WisDOT study. Thus, we noted that a 29" rail height might be a compromise, but further study was needed to ensure that small car underride was not an issue.


For you, the decision will be what level of variation from the TTI approved system you can tolerate. The work Cody did seems to suggest that increased rail heights are better for higher CG vehicles, but the concerns for small cars exist. That said, the TTI study was a paper study that did not test the Yuma system under the TL-2 impact conditions. So neither of the current guidance is founded in a solid crash test. I would think that the 28" height you propose is a minimal variation from the TTI system and may improve the high CG vehicle performance based on what we currently know.

3.    Are the radii listed shown on KDOT's version of the drawing ok? I know from the TTI report it had a radius of 8'-0" (shown as 7.96' on KDOT's version to give a length of 12'-6" for the rail, which is evenly divisible by the post spacing). From the work you and Cody did it looks like radii of 23'-10.5", 47'-9", and 71'-7.5" is also ok. Are radii increments between these radii acceptable? Can those radii be adjusted slightly to give lengths of w-beam rail that are evenly divisible by the typical post spacing (i.e. 23'-10.5" would be ok shown as 25'-0")?

In the TTI report on the Yuma county short-radius, they do not note changing of the radius of the system as one of the acceptable system modifications. This is likely because alteration of the radius may affect capture of the vehicle and energy dissipation. As noted above, Cody's study indicated that larger radii may be an issue as well (however, some of that may have been tied to the height of the rail). Previous recommendations by FHWA have allowed short radius with radii up to 35'. However, that has not been formally verified through a crash test. Thus, it is difficult to recommend larger radii for the Yuma County system. TTI and our research don't seem to suggest that it should be done, but the previous FHWA memo allowed it, so I am sure many states still have the larger radii in their standard. The best option may be to stick with the 8' radius and extend the tangent sides of the system.  If you chose to allow the larger radii, the intermediate values should be acceptable.