Smart questions
Smart answers
Smart people
Join Eng-Tips Forums

Member Login

Remember Me
Forgot Password?
Join Us!

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips now!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

Join Eng-Tips
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

BrianBoen (Civil/Environmental) (OP)
1 May 09 16:39
I have specified a schedule 80 steel bollard with an 8" diameter.  The bollard is 3' above ground and 3' below ground with a 42" deep concrete encasement with minimum 4" cover on all sides of the steel pipe.  

The client would like me to confirm that the bollard will withstand the impact of a 4000 pound vehicle at 30 mph.  How do I convert the 4k pound @ 30 mph impact into a point load or factored load of some sort?

I have the allowable bearing pressure and steel information.  It seems like I have everything that I need except for a load or force that I can use.  I know
KE = 1/2 x m x v^2 and
F = m x a
but I don't have an acceleration and kinetic energy doesn't really help does it.  Can anyone help?
JAE (Structural)
1 May 09 17:57
It can't resist that in my opinion.  The bollard would just lay over.

Vehicle impact resistance is defined for guardrails in AASHTO.  They don't use calculations but rather full scale vehicle impact tests to verify load compliance.

Bollards are more of a psychological barrier vs. a true mechanical barrier.
Bribyk (Mechanical)
1 May 09 18:40
It would depend highly on the vehicle.  Newer vehicles absorb a lot of the energy and reduce your peak forces.  A forklift or something similar wouldn't crumple as much (or at all) and would have higher forces.  If you can assume a newish vehicle you can probable pull up some acceptable g-loads (peak and average accelerations) from the NHTSA for head-on collisions at 30 mph.  I don't know if they test a point load on vehicles though, it's usually a wall-ish type structure.  Then your impacting force can be calculated as F = ma, assuming your bollard does hold up and is rigid.  You could get more complex from there, if required...
CarlB (Civil/Environmental)
1 May 09 18:41
You might check out the documents at

and other "force protection" design guides. Bollards to resist a speeding vehicle will likely need large concrete foundations.
GregLocock (Automotive)
1 May 09 22:39
You can get an upper bound by equating the fully plastic moment of the bollard, *pi/2, to the KE of the vehicle.



Greg Locock

SIG:Please see FAQ731-376: Forum Policies for tips on how to make the best use of Eng-Tips.

dik (Structural)
2 May 09 9:42
I did a bollard for a mining project a couple of weeks back and used a 12" sched 80 stainless steel pipe, conc filled, embedded 6'of 30" dia reinf conc that came to just above grade and then did a sloped surface for runoff.  It's a corrosive environment and use of SS is common.  When the client rep asked how strong it was, I told him that I couldn't find any literature on bollard design and I made it to resist some impact, but that I had essentially made it fairly resistant.  I couldn't find any literature on bollard design other than prescription. I felt that once you deviated from a simple bollard, the added cost of making it stronger was relatively little.  If some equipment actually strikes the bollard, it is possible to shear it off or possibly pull it out.
kslee1000 (Civil/Environmental)
2 May 09 13:23
After studing a few articles on the web, I found the following parameters are useful in develop a simple method to estimate the impact force, with the condition that the driver is aware of the barrier, and has attempted to stop the vehicle, but failed due to inadequate stopping distance. Please comment on the method present below.


Vehicle weight = 4000# (ave. 2300#-3500# for passenger car to SUV)
Deceleration, D = -15 ft/s2 (a ball park number from research, measured from touching the brake to a complete stop. See linked site for more info)
Speed, S = 30 mph = 51 ft/s
Required braking distance for complete stop, B = 59 ft (again, see linked site, also for dist. required for other speed)

Assumptions on Site Condition:

1. Assume barrier is located 15' (can be any) off the drive way. (d = 15')
2. The drive way is level, and frictionless.


Time required for a "reactive" stop:
t(R) = (Vf-Vo)/a = (0-51)/15 = 1.7 s
Time for a "forced" stop:
t(F) = t(R)*(d/B) = 1.7*(15/59) = 0.432 s
Effective deceleration:
a(E) = (Vf-Vo)/t(F) = (0-51)/0.432 = -118 ft/s2
F = m*a(E) = 4000*118/32.2 = 14659# = 15k (rounded)
The load shall be applied at the level of impact.
From here, we can design the barrier and its foundation.

For un-conscientious case (no effort to make stop), since a = 0 at constant speed (Vf just before collision equals the initial speed, Vo), Kinematics couldn't solve the problem. It has resorted to more rigorous energy method.
kslee1000 (Civil/Environmental)
2 May 09 13:34

The derivation of effective deceleration is based on the assumption that the vehicle has crashed onto a barrier with infinite rigidity (no damping), which results in an instant stop (Vf = 0).
JedClampett (Structural)
2 May 09 14:16
What's the bollard protecting?  If it's some bushes or flowers, I'm not sure making hitting it a death sentence for the driver is appropriate.  
Trying to design these type items for a high rate of speed (and 30 mph is high) is nearly impossible.  You make it stronger and the force on it goes up.  And 30 mph head on almost has to be an intentional act.
One of my first designs in my career was to design bollards to protect the frame of a roll up door.  When my boss asked what it was designed for, I responded, "Who are they going to be madder at, us for under-designing the bollard or the guy who couldn't steer his truck?"  Or something like that.  It was a long time ago.
kslee1000 (Civil/Environmental)
2 May 09 14:56
Gorgot the link mentioned above.
GregLocock (Automotive)
2 May 09 19:51

I'm afraid I can't follow your logic, and I KNOW that you can't easily estimate a peak (or even average) force without making some assumptions about the impact process. The link you gave is a bit disturbing in its lack of rigor.



Greg Locock

SIG:Please see FAQ731-376: Forum Policies for tips on how to make the best use of Eng-Tips.

kslee1000 (Civil/Environmental)
2 May 09 21:06

Thanks for commenting. Based on your background (Automotive), I would guess you can provide better sources on similar studies, which may provide easily undersandable clues to solve the long standing puzzle that has bothered many of us in the engineering field.

The method presented is just a quick thinking. It yet to be validated, or discredited, by collective thoughts and sound reasoning. I am eagerly await all constructive critics.  
MaddEngineer (Structural)
2 May 09 21:35
I had a client last year who thought they might need me to design some bollards, I did some limited research, but the project never materialized.

Basically the Department of State has some standards for Anti-Ram Vehicle Barriers and they specify different levels of protection based on a 15,000 lb vehicle traveling at 30, 40 or 50 MPH.  With different stopping distances 0-3ft, 3-20ft, 20-50ft.  (The actual barriers are crash tested per an ASTM Standard) The understanding is that the barrier will deform into the plastic range (or move)and will absorb the impact energy.  The NAVY also had a method based on 10,000 lb vehicle traveling at 15mph or 50mph (Low Security or High Security).  The documents note the Kinetic Energy for each.  For the 10,000 lb vehicle at 15mph the energy will be 75,000 ft-lb (arrived at by using 1/2m V^2 = 75,155 ft-lb).  A true energy analysis could get tricky (you have to look at crumpling of the bumper, bollard, soil, concrete, etc.) but assuming the vehicle will stop in 3ft would give a force of 25,000 lb.  This can easily be taken by an 8in/sch80 pipe.  The bollards will be also most likely 3ft apart?? so at least two (2) would be engaged I would think.  Stopping in 1ft would be 75,000 lb/2 Bollards = 37,500 per bollard.  Might have to do iterations, checking deflection vs. load, however, the stiffness of the bumper/vehicle is an unknown.
kslee1000 (Civil/Environmental)
2 May 09 22:09
That's the problem with energy method, which I tended to avoid simple for the reason - too many variables and unknowns.

For engineering needs, there should be a practical method to approximate the effect. Or more testing should be performed to backup the theory. Try crash GM cars, as they are much cheaper now :)

Note: I am not quite convinced that energy method is correct way to determine impact force. It predicts the collapse mechanism (deformation) by including effect of dampping, but the question remains - the force at the split second when two rigid body colide, tome, damping is part of thereafter, which does not have relevance in determine the force at collision. Again, I could be deadly wrong on this.   
GregLocock (Automotive)
3 May 09 1:35
OK, the KE=plastic energy in bollard approach gives an upper limit, that is, if the KE of the vehicle is less than that required to bend the bollard out of the way then by definition the bollard must be still in the way.

But, for real vehicles that is almost absolutely useless, since vastly more of the plastic deformation will occur in the vehicle, not the bollard.

Passenger cars/SUVs aren't really designed to cope well with poles, frontally, which will tend to cheesewire through the structure. That's why Armco fencing is designed the way it is - the contact stiffness is low and there are several weak mechanisms for absorbing energy.

Your best bet would be to study some pictures of poles cheesewiring though cars, to get an estimate of the crush distance involved for a given speed. You can then turn that into an average force, and double it to get a sort of peak force, and double it again, for luck.

At 30 mph 13.4 m/s I guess that you'd see 0.5-1m of intrusion, so for a 2000 kg vehicle the 'average' force by this method is 1/2*m*v^2/s=180-360 kN

So, that is 9 to 18g, which is in the ballpark for a crash pulse.

So if you double it and double it again I'd design the bollard for 1440 kN, exerted 0.6m off the ground (the height of the engine).

Note that this will damage the car and its occupants far more than a deformable barrier. This is a pretty severe event.

Also you need to make sure the car doesn't jump over the bollard, that's why an elastic solution is better. If the bollard lays over then all bets are off.



Greg Locock

SIG:Please see FAQ731-376: Forum Policies for tips on how to make the best use of Eng-Tips.

kslee1000 (Civil/Environmental)
3 May 09 10:22
My intent wasn't limited to bollard, it's for general applications.

Using energy method, assume linear motion with known structural rigidty of system (vehicle and barrier, replaced by an equivalent spring), here is the foundamental relationship between parameters m (mass), K (equivalent spring constant), x (displacement), V (final velocity), a (acceleration/deceleration):

U (work/energy) = K*x^2/2 = m*V^2/2 = m*a*x, (F*x)
From the relationship, we can setup two equations to solve 2 unknowns - "x" & "a".
The result is, a = V*(K/m)^2/2
K is either from test, or could be calculated from known data on barrier support and the vehicle.
V (at time of impact) can be safely assumed as equal to the given speed (ie. 30 mph for the case under discussion)

The the impact force is simple, F = m*a.  
GregLocock (Automotive)
3 May 09 20:26
Yes, if you know K then you can derive a peak force quite easily.

But the effective k in a highly non linear event is not obvious, and in my opinion is almost meaningless.

Hey, here's an article we need to read

but its 34 bucks, so I guess the OP gets to buy it!


Greg Locock

SIG:Please see FAQ731-376: Forum Policies for tips on how to make the best use of Eng-Tips.

GregLocock (Automotive)
3 May 09 21:22
A quick look at the search results for pole impact at indicates that there isa  fair amount of literature on this.

I would come back to a previous point, cheesewiring through a vehicle could inititiate multi million dollar lawsuits, is your structure really that valuable?


Greg Locock

SIG:Please see FAQ731-376: Forum Policies for tips on how to make the best use of Eng-Tips.

kslee1000 (Civil/Environmental)
3 May 09 22:38

Bollard has never intended to completely stop a vehicle. It generally serves as a warning by the vehicle entrance, and as a back-in guide. The 30 mph request was not reasonable, no one would try to achieve that without given the owner the warning as you mentioned. However, I was tring to take upon this opportunity to find a practical method to calculate the impact force, which has bothered many of us, and has broad, though rare, applications. Thanks for keep coming back with valuable informations.
bridgebuster (Civil)
4 May 09 8:37
Before you do anything, look at the attachment. It has some very important imformation on bollard design.

You can try a down and dirty approach by assuming the bollards are cantilevered beams or piles with lateral loads. The point of fixity would be somewhere within the concrete foundation. Check the foundation to see if it has enough breakout strength and check the depth to see if it has enough passive resistance against overturning.

Where I work the building is surrounded by bollards. They were put in five years ago; just before the Republican convention as a security measure against a truck bomb. They are 8"-dia pipes filled with concrete; the bollard spacing is 5'-0; they extend 2'-6" above grade. The footing is 18-inches wide. I don't recall how deep the pipes are embedded. The top of the foundation is about 6" below grade; I think it's 3- deep.

On a bridge project we put in bollards as a pedestrian safety measure against a bus mounting the sidewalk. The bollards are 6"-dia Schedule 80 pipe; 3'above grade. They're mounted to a 3' wide x 2' deep footing, 10" below grade using 8 - 1" dia anchor bolts with 10" embeddment.  
kslee1000 (Civil/Environmental)
4 May 09 9:54

When you say,

"Yes, if you know K then you can derive a peak force quite easily.
But the effective k in a highly non linear event is not obvious, and in my opinion is almost meaningless."

I would like to remind you that, while I agree with you that the whole event (car crash) is likely ends up in the state that renders K meaningless, it (total energy involved) is not really in our interest. For most of engineering problems, we don't need to get everything "exact" as scientist does, but to obtain a "reasonable accurate" solution through a manageable method that is in line with the theories behind it. That (manageable method) is what we are looking for.

How to get K? For structural support system, it is tedious to cal, but solvable. For vehicle, it is much murky, since I don't think there is any study been done for that. Therefore, I would suggest to treat the car as a rigid body, and the structural support system as the spring. By doing so, though it is not the "exact" condition, but rather conservative (to assume no damping effect from the deformation of the vehicle).   
BrianBoen (Civil/Environmental) (OP)
4 May 09 11:56
I must say thank you to all that posted.  Your information is valuable.  

There were a few questions that I saw:  First, the bollards are going to be in front of a Federal Courthouse.  So the strength of these bollards must meet criteria as stated above.

I was wondering how you came up with 75,155 ft-lb of energy for a 10000# vehicle @ 15mph using the 1/2 * m * v^2 formula.  

MaddEngineer was correct in mentioning the stopping distance requirement.  The ASTM standards vary from one of three options: 0-3ft, 3-20ft and 20-50ft.  I forgot to mention that in my original post.  The requirement that needs to be met is 3-20 seeing as the building is about 20 ft set back from the bollard.

I don't think the cheesewiring would be an issue with that being said.  We are not designing to completely demolish the car and stop it on a dime rather we want to slow it down and not allow it to travel any further than 20 ft beyond the bollard.

Does that new requirement shed any light on your theories?  It seems that kslee1000 is on the right track and the 20ft stopping distance requirement should be usable in the deceleration equations you provided.

JedClampett (Structural)
4 May 09 13:30
This is the kind of issue that comes up in Nuclear Generating Station design.  There might be some literature available there.  However, they aren't worried at all about aesthetics, so there's a lot of wire with multiple layers of barriers.  
I'd use a group of staggered bollards with a 4'-0" spacing and another row about 4'-0" behind them such that there's no way you can enter without hitting one bollard square and then two behind it or vice versa.  If anyone gets through the first set, the second set will certainly stop them.  And it shouldn't impede wheelchairs or pedestrians.
Helpful Member!  Larryhd2 (Structural)
4 May 09 13:50
Here is an article you might want to review...
kslee1000 (Civil/Environmental)
4 May 09 15:00
JED's suggestion is worth to think about. Or how about concrete block barriers?

By throw in the allowrance on stopping distance made this problem even trickier, because now it involves inelastic behaviors of both subjects (car & barrier), and failure mechanism of the support system. Not an easy task for computation, hope something readily available is out there somewhere.  
BrianBoen (Civil/Environmental) (OP)
4 May 09 17:27
It's not a calculated solution that I found but it still works.  Looks like I'll have to increase the depth of the steel pipe and the size of the footing a little bit.

See the attachment for the detail.  It has been tested to withstand the impact of 4500# @ 30 mph.  It was found in FM 5-114 Appendix A and in the US Department of Transportation document titled "Transit Security Design Considerations" (FTA-TRI-MA-26 7085-05) in Appendix E.
Larryhd2 (Structural)
4 May 09 17:35
The article I posted will pretty much answer your questions regarding the load imparted to a bollard.

Assuming a zero deflection at the bollard I came up with about 61kips to the bollard.

You would need to do a trial and error with bollard deflections figured in to dial in a more accurate loading..  making sure the bollard stays within the elastic range...
BrianBoen (Civil/Environmental) (OP)
4 May 09 18:00
Do you mind explaining how you got 61kips?
kslee1000 (Civil/Environmental)
4 May 09 18:34

Very interesting site.
Below is the link for persons interested in Transit system. I guess there are quite a few free reports for dowload.
Larryhd2 (Structural)
4 May 09 18:59
Honestly, I looked at the chart in Figure 5 in the article and scaled down for a 4000# vehicle...

Using equation 5 given, I come up with a vehicle crush of (30^.5)/3 = 1.83'

Using equation 2:
F=mv^2/(2*(1.83+0)) = (4000/32.2)((30*1.4667)^2)/(2*1.83)

Assuming the bollard moves (deflects) 6", the load goes down to mv^2/(2*(1.83+.5)) = 51,600#.....

It seems you would then need to decide on some bumper height to apply the load, determine bollard deflection under the applied force, and iterate...  
Bribyk (Mechanical)
4 May 09 20:01
How about putting some stairs in instead?  
BrianBoen (Civil/Environmental) (OP)
4 May 09 20:04
wow, I feel stupid.  I wasn't dividing the weight by 32.2 to get mass.  I kept coming up with this astronomically big number. haha.  Thanks.

I believe about 2 to 2.5 feet would be an appropriate location to apply the load.   
BrianBoen (Civil/Environmental) (OP)
4 May 09 20:08
Stairs are a good idea for a new site.  This is an existing site and unfortunately is not a possibility.
Larryhd2 (Structural)
4 May 09 20:09
Good ole slugs will do it every time...
JStephen (Mechanical)
4 May 09 21:49
This won't help with the design problem, but is a cool video.  An Anti-Ram wall at work- takes a few minutes, but about halfway through, they show a loaded dump truck smacking into it.  Impressive.
kslee1000 (Civil/Environmental)
5 May 09 11:46

I have to praise Larry for the paper he provided. It seems reasonable yet simple to use. How is the project going?
patprimmer (Publican)
5 May 09 21:14
If you want to stop an aggressor, look to WW11 tank trap designs or designs used in ancient times for foot soldiers to ward off calvery with strong sharp steel spikes firmly anchored and pointing at the approaching vehicle.

To prevent entry and give maximum protection to the driver, look at drag race catch nets where a woven wire mesh is used and mounted in a way to yield quite a distance it absorbs the energy and distributes it over the entire front f the vehicle.

Really solid bollards tend to cut cars in two up to at least the engine, or if less solid, knock down and then the car gets airborne as it goes over the top of the sloping post. If the concrete the post is embedded into has insufficient bearing area in the soil it will just knock the post over. I think there is a real need to think the real objectives and possible results and consequences through. The bollard in a relatively small amount of concrete in variable ground with various vehicles at various angle has a huge number of variables.


See FAQ731-376: Forum Policies for tips on use of eng-tips by professional engineers &
for site rules

kslee1000 (Civil/Environmental)
5 May 09 21:44
Unfortunately, the request for barriers set-up will keep rising, as long as we live under constant fear. As Brian pointed out, he was working on behalf of the Fed, I don't think the barrier is for you and me, but... So the criteria for speed, 30 mph, is way high than normal used for warehouse, or shipping-receiving terminal.
MaddEngineer (Structural)
6 May 09 9:28
Sorry Brian, for not responding sooner, but looks like you figured it out (m = Weight/32.2)... Also I didn't realize you noted the design criteria (4000lb @ 30mph) - (Must have had a brain cramp, was cranking calcs out on a Saturday night and I have two twin baby boys at home)... Larry great article that Equation 5 is the motherload.  Looks like Brian has enough info to rationalize his design if he distributes the load to two bollards he will remain in the elastic range.  However, I guess the problem still remains for mass-velocity relationships that put you into the plastic range for the bollard(s).  I was thinking of modeling the bollard as a rotational spring that would rotate at the plastic hinge. I found a paper online discussing plastic deformation of a cantilever under tip impact but its a little heavy and I haven't really took a hard look at it for relevance (if anyone wants a stab check it out).  Also does anyone think some (or alot of)energy will be initially absorbed by the foundation since theoretically the foundation should slide until it is stopped by something, either the soil or the roadway and then will it engage the bollard in bending (albeit this may be a very small distance).  I saw evidence of this at the drive thru a McDonald's, bent bollards and cracked/crushed pavement.  You could use the EA of the footing (a large number) plus the k of the soil... Any thoughts
kslee1000 (Civil/Environmental)
6 May 09 10:08

Glad to see others to join in.

I thought about rotational spring analogy as well. There are two scenarios on how the system behaves:

1. Barrier on massive concrete (infinite rigid) block below  grade: The barrier acts more or less like cantilever beam.
2. Barrier on foundation with intermediate rigidity, like in the shaft filled with concrete: the barrier may work with its foundation as an integral unit, the failure mode is linear in nature (the system rotates like the needle arm in clock). At the very moment, the impact force has caused the soil wedge in front of the footing to reach passive state, thus the soil resistance is lost.

If a sliding distance is allowed after initial impact, one can manipulate to design a more flexible system, which is allowed to fail, but effects in slow down the vehicle. The vehicle would then come to a complete stop by friction within the pre-set distance.

The concept is simple, which can be expressed as:

Ff = Fi - R - N = 0

Fi can be tetermined by displacement obtained using spring analogy; R is the resistance force could be estimated from the subgrade condition; N is the friction force.

How the above matches your thinking?   
MaddEngineer (Structural)
6 May 09 12:14

Scenario 2 seems to be inline with what I'm thinking in terms of behavior, but its been a long time since I took a dynamics class so I'm a few pages behind.  But I think you stated it eloquently, but, this is a great exercise for something as you stated will be coming up more and more often.  

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Back To Forum

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close