Pipe bollard dynamics

[RacingAZ]

I have this pipe bollard that is rated and crash tested per ASTM F2656 as M50 / P1. What that means is that a medium duty truck (15,000 lbs.) running at 50MPH upon impact with the bollard, stopped no more than 3.3ft. beyond the bollard line. Per the reference ASTM standard above, this would have produced a kinetic energy of 1250 ft-kips. I don’t have the specific information about the condition of the bollard after impact.

Here are the details of the bollard assembly. It’s a 10” dia. sched. 80 steel pipe with additional ¼” thick plate welded inside (parallel to the direction of the impact) and in-filled with grout. This bollard is embedded in a 32” wide x 48” deep continuous concrete grade beam. Bollards are spaced at 48” o.c. but only one bollard is impacted in the testing.

I’m trying to replicate the results of the testing by running the numbers. Looking at the steel bollard capacity (ignoring the grout in-fill for now – I don’t think that the grout adds a significant number to the overall strength of the bollard). The bollard is not even close to taking the full amount of force generated by the 15,000lbs. truck at 50MPH.

This got me thinking about the force dynamics involved. The bollard isn’t really taking the full energy of the truck upon impact. I’m assuming that some were lost due to friction resistance of the tires against the pavement. The grout inside the pipe also takes some of the energy. Also, there is an instantaneous transfer of energy occurring from the bollard to concrete grade beam then to the soil at the moment of impact while the bollard is yielding.

Are these assumptions valid? Other assumptions I might have missed? Anyone here has any experience designing bollards per ASTM F2656 or SD-STD-02.01?

Also, how do you quantify each assumption? I do have a vague idea but would like to find out what your thoughts are. I’ve seen some discussions here about bollards but have not found the definitive answers to what I’m looking for.

Sorry for the long post and thanks for anything that you can add to this discussion.

 

[demayeng]

I don't have experience in this although I am interested in the outcome... I have to design a similar thing

I would assume that the bollard would be taking most of the energy - the friction of the tyres over the stopping distance would be negligible, and energy is only transferred into the soil if deformation occurs??

The force applied to the bollard can be calculated by dividing the KE before impact by the stopping distance. So you would get 378.8 kips on the bollard. I guess you could look at the plastic modulus and use a bending moment at the base of the bollard to see if it can handle this??
That's all i have at the moment!
I'll be interested to hear anyone else's input..

 

[RacingAZ]

That's pretty much what I have but applying 378kips to the bollard, say 27"-30" above finished grade, the bollard would be 6-7 times overstressed. So in theory, there is no way the bollard can take the full KE.

I have now seen a couple of videos of the crash testing and the bollard didn't even bend upon impact of the truck. It sliced thru the center of the truck and the front of truck (what's left of it anyway) protrudes a certain distance from the bollard.

You would think that the bollard would get sheared off or bent down, but that is not the case.

 

[DaveAtkins]

Doesn't the truck absorb a portion of the kinetic energy as well? Or does it suffer no damage?

DaveAtkins

 

[csd72]

I believe that dave has the answer.

The truck is much softer than the bollard.

 

[RacingAZ]

I agree. Most of the KE's absorbed by the truck as the bollard didn't suffer any damage. With the bollard being rigid, the bollard impact force is the same amount absorbed by the truck?

See attached article for a discussion of energy-based design method.

Based on that article, it confirms that the bollard is subjected to 378.8 kips of force as demayeng noted. What I'm having a hard time grasping is that the calculated theoretical bollard capacity, using the listed yield and tensile strengths of the pipe (A53 type E) is nowhere near the 378.8kips.

The only unaccounted element is the grout in-fill inside the pipe. Am I to believe that the grout will make up the difference or the pipe's actual strength is way more than what is listed?

 

[Compositepro]

The major oversight in the analysis so far is that the truck is a relatively soft object and the stopping distance of the rear part of the truck is much greater than the front part. If it were solid steel (which, in effect, has been assumed so far) then a far smaller mass would be capable of shearing the bollard.

 

[demayeng]

If the bollard doesn't bend, then the only thing giving it a 3.3 ft stopping distance is the deformation of the truck, which means it is the rear of the truck that has the 3.3 ft stopping distance.

The force required to create this deformation is the same as the equal and opposite force applied to the bollard.

I guess the grout fill is completely confined, meaning it would have high strength. It would also aid to prevent the pipe from buckling. So the grout-filled pipe could act as a very efficient composite steel-concrete column

This paper says they got an increase in bending strength of 25% by concrete filling:

http://flyashbricksinfo.com/constru...crete-filled-steel-hollow-sections-beams.html

Not enough to explain our situation though..

 

[RacingAZ]

compo, the analysis follows the actual crash testing behavior of the truck and bollard. In the actual crash testing, the front of the truck gets crumpled or deformed upon impact against the bollard at the same time that it slices thru the front of the truck.

What I'm actually trying to do is attempting to correlate the crash testing behavior against the generally accepted scientific formula for determining the theoretical strength or capacity of the bollard and it's not even close. Why the wide discrepancy? That is the question that I'm trying to solve.

 

[paddingtongreen]

You cannot assess the strength in those terms unless you also specify the spring rate of the truck's collapse. Rather than the rear of the truck, it should be the impact and final positions of the center of gravity that are measured.

I was told, early in my career that this is what has been found to work in practice, don't question it.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[azcats]

RacingAZ - What are you using to calculate the capacity of the bollard?

 

[Compositepro]

"stopped no more than 3.3ft. beyond the bollard line". This would mean that the front of the truck does not pass that line. The back of the truck stops in a much greater distance.

 

[chicopee]

Why not forget the energy absorption by the truck, that way you would have a system with a greater factor of safety but costlier.

 

[JStephen]

Also consider that you are accelerating the bollard and some of the soil or concrete around it, so you can't assume forces are in equilibrium to calculate bending moments.

As to the truck travel- does it mean the truck never goes past that point, or rebounds to less than that point?

 

[demayeng]

The 3.3 ft for the rear (or centre of gravity) will only be increased by the amount the front of the truck crumples, although it sounds like it mostly cuts through. At the end of the day it doesn't matter though - we know roughly what the stopping distance is, whether it is 3.3 or 5 or 6 ft won't affect the calculation by the amount the bollard strength is out...


The spring energies of the bollard and the footing/soil that JStephen refers to must act (a lot!) to reduce the overall force experienced by the bollard. The bollard might not be too hard to work out, but the footing..