I am sending you an email response that I had sent in response to a similar question. I hope that this new information will help answer your question on unsupported guardrail spans greater than 6 ft - 3 in. Please note that the question raised to us previously related to long span guardrail systems in general and the associated nesting requirements for the W-beam rail adjacent to the unsupported span. Your question was very similar in that you desired comment on a strong-post, W-beam guardrail system where one post was left out. For a quick answer, you may want to jump to the last paragraph. In summary, with current W-beam systems, we believe that you will need to nest the W-beam guardrail over the unsupported length and carrying the nesting into adjacent regions. At this time, the length of this nested rail for spans less than 25 ft is undetermined nor optimized for impacts with pickup trucks at the TL-3 conditions. In the future, we believe that the use of the MGS, in combination with long-span systems, may actually allow us to significantly reduce and/or eliminate the need for nested W-beam rail.
Per my previous email response:
Recently, you had asked me to consider what reduced nested guardrail lengths would be acceptable for long-span guardrail systems placed over shorter culverts (i.e., with only one post (12.50-ft span) or two posts (18.75-ft span) left out). Currently, we use 100 ft of nested rail for the 25-ft long-span system which resulted in 37.50 ft of nested rail on each side of the unsupported span. Obviously, shorter nested rail lengths will be acceptable when the unsupported span is reduced from 25 ft to either 12.50 or 18.75 ft. Three general approaches could be utilized for considering the shorter span alternatives. They are as follows: Option 1 - A very conservative approach, which results in significant overkill in the design, would use 100 ft of nested rail for all long-span systems (i.e., 25 ft or shorter spans) in the absence of any new research (crash testing or computer simulation modeling). Personally, I would not recommend this approach since excess nested rail would be used for systems not requiring it. However, some states may choose to be very conservative and employ this solution. For clarity, this approach would consist of the following: 25.00-ft span: 100.00-ft nested rail 18.75-ft span: 100.00-ft nested rail 12.50-ft span: 100.00-ft nested rail Option 2 - A moderately conservative approach would use the same length of nested rail on each side of the unsupported span or 37.50 ft. For shorter spans, this would result in a total nested rail length equal to 75.00 ft plus the length of the unsupported span. Prior MwRSF testing showed that the 37.50 ft of nested rail on each side of the span would be adequate for reducing and/or eliminating the potential for rail rupture when used in combination with a 25-ft span. As such, I believe that this same nested requirement would also be more than adequate for the 12.50 and 18.75-ft spans. For clarity, this approach would consist of the following: 25.00-ft span: 100.00-ft nested rail 18.75-ft span: 93.75-ft nested rail 12.50-ft span: 87.50-ft nested rail Option 3 - A slightly more aggressive approach would use a reduced nested rail requirement on each side as shorter unsupported rail spans are encountered since it is recognized that dynamic deflections and critical rail stresses will likely be contained within a shorter region. In the absence of a BARRIER VII computer simulation effort or any compliance testing, I am personally unwilling to recommend this option. Please note that I am not stating that this option, or even one more aggressive yet, will not perform in an acceptable manner. I am just unwilling to recommend such a leap without further validation/verification. However, I will provide it for clarity below: 25.00-ft span: 100.00-ft nested rail (37.50 ft of nested rail on each side) 18.75-ft span: 81.25-ft nested rail (31.25 ft of nested rail on each side) 12.50-ft span: 62.50-ft nested rail (25.00 ft of nested rail on each side) Finally, it should be noted that the relocation of the rail splice and/or implementation of the MGS guardrail system for culvert applications may help to reduce the nesting requirements from what is shown in all three options above. However, those nesting reductions can only be verified with full-scale crash testing according to the NCHRP Report No. 350 criteria, computer simulation modeling, and/or combinations thereof. It should be noted that several Pooled Fund States have inquired about this same topic in the past, that of which must be resolved in a future funded research study. If you have any questions regarding this matter, please feel free to call or email me at your convenience. Dean would also concur that options 1 and 2 are too conservative and even the final option (option 3) may be actually over designed.
The MwRSF Pooled Fund States funded several studies to evaluate the safety of the MGS with a missing post. Results of those studies may be found in the following links:
https://mwrsf.unl.edu/reportResult.php?reportId=317
https://mwrsf.unl.edu/reportResult.php?reportId=357
https://mwrsf.unl.edu/reportResult.php?reportId=395
These studies included systems with and without the use of a curb. When the MGS was evaluated according to MASH2016 TL-3 impact conditions on level terrain, the system performed as desired with a missing post, although some rail tearing was observed. During MASH test designation 3-10 evaluation of the MGS installed 6 in. behind the front face of a 6-in. tall, AASHTO Type B curb, the rail ruptured and the vehicle penetrated through the system. Subsequently, 37 ft – 6 in. of rail was nested centered at the omitted post, and test no. MGSCO-2 was determined to be successful. Subsequent evaluation according to MASH test designation 3-11 of the nested rail configuration with omitted post was also determined to be successful.
Further guidance related to implementation of the missing post configuration is presented in the research studies above. Briefly:
Some parts of this site work best with JavaScript enabled.