soil pressure on irregular shape

[dison]

I was just interested in how other engineers might determine the loads on a wingwall retaining wall shaped as shown below. I believe that the way I did it was fairly conservative, but would like to hear ideas from people just looking at it for the first time. The wingwall is parallel to a railroad track and will have active soil pressure as well as boussinesq surcharge. The embankment outside the wall might not be above the bottom of the wall, so I cannot consider any passive resistance. It cantilevers out from a bolted connection to the side of the bridge abutment.

_________________________
/ !
/ !
/ o !
/ !
/ !
! o o _
! _
! o o _
! _
! o o _
!___________ ~dison

 

[kieran1]

I would work out the horizontal pressure on the wall due to the soil pressure and surcharge for the depth of wall and design the wing wall as a cantilever with a triangular distributed load which is equal to the horizontal pressure at a given depth
the wing wall is cantilevered off the abutment so the connection would be fairly substantial Kieran Coyle
BEng(hons),CEng, M.I.Struct.E, M.I.E.I.

 

[ishvaaag]

You are taking unto account soil (assumed damp if possible so) and one way or another the effect of the loads above, so in general your approach is correct. Only if the wingwalls are too close the reduction of pressure coming from a cut active wedge should be accounted and not doing is reasonably conservative. Then the stresses are determined according to the boundaries and structure.

 

[dison]

How would you distribute the pressure over the sloping portion at the bottom of the wall and resolve it into design forces? Would you try to do it exactly (3D pressure - integrate?) or would you do a quick & conservative estimate based on some percentage of the total load on a totally rectangular wall? ~dison

 

[ishvaaag]

I presently would use a 3D design quite for sure, since modeling plates (3D faces in Autocad) and then importing to say RISA 3D is easy and quick. The loads I would calculate as if the wingwall, yes, was a complete rectangular wall, reading the values of pressures at the depth, to load the proper plate elements.

 

[Qshake]

It seems that the ultimate goal here is to design the bolted connection and to have a sturdy plate capable of transmitting that force (soil pressure) to the connection. And I don't see why it is necessary to anything more than to the load the wall with the appropriate limits of the soil diagram that would result from elementary soil mechanics and the appropriate influence of other loads (surcharge). Following that find the resultant of that force assuming the wall as a cantilever and determine the force and moment on the connection. This moment can then be used to design the plate thickness as well - or design the plate based on the capacity of the connection or vice-versa. The latter, I note, is due to the fact that most railroads (assuming this is a wingwall application for a railroad - not many DOT's doing this!) would rather have something a little stout to ward off the inevitable section loss from environmental concerns.

I don't see the need to work with 3D or finite elements. This is a statics problem to determine the load and strength of materials to size the connection and plate thickness.

And yes, I do enjoy using finite elements quite a bit - when the application is truly warranted.

Good Luck.

 

[dison]

Thanks for all the replies.

I agree with Qshake that this probably doesn't warrant the use of finite elements and elementary soil mechanics can be used to determine the appropriate loads. It seems that I forgot to mention the material being used for this member. Ihe wall isn't a steel plate, it is actually a precast concrete piece. To save some time, I figured the soil pressure diagram and surcharge on a rectangular section that was the full depth of the wingwall, divided the wall into two sections (above and below where the sloping bottom starts), and applied the total load above the divide and 2/3 of the load below the divide. This yielded reasonable values to design with and so I did not recalculate the loads exactly. This should be conservative since the max pressure is at the base of the wall and at that point the wall is actually sloping toward lower pressure. ~dison

 

[Qshake]

Ah, the precast. Its amazing what the railroads will do with precast. This is for a railroad structure isn't.

I watched a crew of about five guys (who knows if they even belonged to a union) put together a new railroad bridge not too far from my house. The abutments were all precast pieces (they sat a big block down on the old abutment and affixed a precast backwall to it. They drove cylindrical pile down between the track and placed a precast pile cap beam ontop and one day when the train traffic was light they pulled the old superstructure out and placed a new one (actually looked like a dupont overhaul as the newly painted girders were riveted!). All that before the next train coming down the track. Amazing!

 

[RiBeneke]

In this application, where the backfill is subject to vibration from the trains, would you calculate your pressures using Ko (at rest pressure estimate) than Ka (active pressure). I think I would use Ko.

 

[ishvaaag]

Good precision, Ribeneke...

And if the dynamic loads of the locomotives are akin in impact to vibrating rollers the pressures may go instantly amazing ... experimental impact factors must be covering this...but if only moderate ones are used in bridges tehmselves to apply bigger in approaches would seem strange.

 

[dison]

RiBeneke & ishvaaag:

In my experience, impact has not been applied to railroad surcharge loads for design of retaining walls and abutments. I don't know if the reason is due to the damping effect of soil or factors of safety applied in design. In this case, the wall is 14 feet from centerline of the track and the surcharge load at that distance is minimized. If the wall were much closer, that would be a much more important factor in design.

Would you use at rest pressure with this type of cantilevered system and bolted connection? My instinct would say that under soil pressure, the wall would be flexible enough to go active, but my instinct is certainly not always correct

Qshake:

Yes, it is amazing what the railroads will do with precast. Whatever it takes to keep the trains running! It makes for efficient construction (assuming that no problems creep up during track outages). The bridge construction sequence that you mention is very common. Where did the term "dupont overhaul" come from? I have never heard that one before. Are you saying that no one is building riveted girders these days? ~dison

 

[ishvaaag]

The active soil pressure is used for cantilever walls even with a heels within the soil in the understanding that the tip outwards required movement will be anyway (and maybe anytime, through support of the wall adaptation) available.

Contrarily the at rest pressure is understood to develop against nonyielding walls, such a rigid buried box.

One may argue that once the wingwalls have taken the deformation demanded by the pressure, they become nonyielding, but this could also may be said of the cantilever walls with heels in the backfill and the most common use is that of active pressure, irrespectively of the actual upper tip lateral deformation. Furthermore, as in such case, in wingwalls the confinement is as commonly understood for retaining cantilever walls imperfect, and this will signal the use of merely the active pressure as enough.

Other thing is that one feels the pressures being applied be less than required on engineering judgement (for example, on the namend lack of account of impact factors that care for the dynamic load) and then in compensation choose to -for example- use the at rest pressure.

Technically the check would be if the required displacement to develop active pressure is met.

 

[Qshake]

`dison,

Perhaps somewhere riveted construction is being carried out. But not here...

Anyway, as a new graduate I worked in the construction field and became very familar with the "dupont overhaul" terminology! Just because it looks and smells brand new that is no reason to expect it to run any different than the 30+ year old hunk of junk it was before the coating! Of course, with respect to the big plate girders I have to commend the railroad for rummaging around and making the find - very economical!!