Theory question (fillet weld loaded perpendicularly to faying surface for a temporary condition)

[jrw501]

Suppose you have a steel box sitting on a steel plate with a force applied at the top perpendicular to the box. The box is heavy enough that the load doesn't cause overturning and the friction force is large enough to prevent sliding.

Now suppose you fillet weld the sides of that box to the plate. Should those welds be designed for the shear from the applied force + bending about the weld group (caused by the load applied at some height)? Or is it fair to say that the welds won’t be engaged unless friction is overcome or tipping is expected to occur?

The thing I’m actually looking at is temporarily restraining a bridge bearing to resist a longitudinal force from a train (the longitudinal force is going through a pin at the top of the bearing assembly). The contractor wants to add some short welds on three sides of the base that they have access to just to make sure it doesn’t move. I don’t think it’s a problem since it’s not going to worsen the condition (worst case the welds fail), but my question is more to enhance my understanding.

Sorry if this question is very basic. Intuitively I’d want to design the weld at minimum for the difference between the friction force and longitudinal force (if the friction force were smaller). I’m not sure how to deal with the (force*height)/(weld group section modulus) force though.

Thanks!

 

[desertfox]

Hi jrw501

Yes the welds should be designed for shear and bending see this link:-

http://www.roymech.co.uk/Useful_Tables/Form/Weld_strength.html

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[TehMightyEngineer]

In theory I believe the box welds can be designed for no force. The stiffer support is the friction and self-weight preventing overturning. Imagine you put the load on the box and then welded it in the loaded condition; there will be no change in the load path. If the box was flexible enough I could see some stresses getting to the welds just due to the deformation restraint provided by the welds but for theoretical discussion I'm assuming no significant internal deformation.

Now, the real world is different. In your real situation of a heavy vehicle load there will be vibration, impact loading, unanticipated friction and self-weight, etc. This means that friction becomes less reliable as the sole means of restraint. In addition, real structures do flex and thus the welds may provide a stiffer load path and thus if they are installed may take the load vs the friction. Finally, safety factors and all that means that the friction and self-weight countering overturning must be reduced so therefore you may no longer have sufficient restraint with those alone.

When you design the weld I would design it for the full force. If/when the bearing does slip the full load will be transferred to the weld; trying to share the load with the friction wont work.

All said, I'd say your theory is correct and your contractor is correct.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[jrw501]

Thanks for the responses and the reference!

So the confusing thing to me is that it seems like by adding a weld problems are just being created because the welds are going to fail if they actually see that moment -- but I'm not sure that they're problems we're going to see in the real world.

Specifically in my case the vertical loads (from gravity and live load+impact) are about 675k (AREMA uses ASD so these are unfactored loads). The longitudinal force from a train traction is about 130k applied at about 3'9" from the base of the bearing.

Even assuming a relatively low steel on steel friction coefficient the friction force is going to be about as big as the traction force or greater (and in the dry condition, much much greater). Overturning about the edges won't be an issue (the base is about 3'4" long x 4'0" wide). So in its current condition this thing is just going to stay put.

But now, if I add welds on 3 sides (say one on each side of the length 40" each and one over the width of 48") to take the longitudinal force -- even if I use a 1/2" weld -- resolving that moment of about (130k*3.75')=~488 k-ft is going to result in significant overstressing of the weld group (and the contractor only wants to use something like 5/16" welds on 3 sides that are only about 6" long each).

I'm not thrilled about relying on friction to prevent movement, so I was thinking maybe it's fine to design the welds just for the full longitudinal force, but I'm having trouble explaining away the bending aspect. And if I include bending it just doesn't seem like it's possible to make work.

 

[TehMightyEngineer]

Well, overturning resistance from gravity loads is a reliable force. Gravity doesn't go away, so other than underestimating the contribution of the gravity loads or under estimating the overturning forces you should be able to ignore that force for weld design (with a rational safety factor).

The other aspect to consider is that welds are ductile and actually have a decent amount of capacity beyond their strength. I've overloaded welds in tension (load testing a lifting beam) and they will start fracturing at stress concentrations (roots, notches, etc.) but you will have additional strength past what would be called "failure" of the weld. So, it's not like the welds would do nothing in an overload situation but I wouldn't call them "safe" in an overload situation.

For added safety I'd say you're on the right track. Size them to resist the sliding force and call it a day.

Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[jrw501]

That makes sense, thanks a bunch!