Severely deformed saddle

[KLee777]

Well, this is a new one. We received a new vessel and found during installation that the saddle baseplate was severely distorted. (See attached photo.) The gap is as much as 1/2" at the edges of the baseplate.

The OEM is suggesting we shim this, but I disagree that any amount of shimming would provide a fully supported baseplate. My colleague says he isn't concerned about this and that fixing it wouldn't be worthwhile. I disagree, as the forces aren't being transferred to the skirt the way it was originally designed. Knowing the unit this is going into, it's not a high pressure unit and the operating liquid level is very likely minimal. I don't see much "flattening" going on even when in service.

Would appreciate your thoughts on what you would do.

 

[dhengr]

KLee777:
It seems to be doing fine right now, so what’s the problem? There is a general concave up shape in the base pl., but there appears to be a gap, on the camera side (end) btwn. the support beam and the saddle base pl. as well. What does the base pl. at the other end look like, or is this just an isolated instance of distortion and poor fabrication? I’d leave it alone and shim it. You’ll do more damage (to the vessel and/or the saddle) by trying to fix it than it’s worth. Loosen the two pairs of bolts and measure the gap on either side of the bolts and then again 3" further away from the bolts. Have a machine shop cut the measured slopes across some pieces of 3" bar stock, by a length a little shorter than the saddle base pl. width. Put these in place and tighten down the bolts again. Take another 3-4" piece of bar stock and have them cut a tapered shim which should be driven in under the center stiffener pl. on the saddle. And call it a day. Of course, the OEM should pay for this or do it. A perfectly uniform loading (fully supported baseplate) on these saddle bases is really just a dream or a wish. You might assume the loading on one pair of those shims out at the bolts, (.25 or .3)(total vessel weight)(?); then check this load against the saddle web bearing and crippling strengths, over the 6-8" width of the two shims. Take a look at the same thing at the tapered center shim, as it relates to bearing and crippling of the stiffener pl. If this checks you should be o.k.

 

[jtseng123]

This rocking chair is not uncommon. This is due to welding/heating as I was told by a fabricator. I saw one proejct a lot of diffrent thickness of shims were shown for a horizontal drum. May be for that purpose.

 

[JStephen]

If it was on concrete, you'd shim and grout it. Can that be done for steel on steel?
Looks bad. But looks like if the saddle "failed", it would be forced into proper contact, so it's hard to see how that's really a problem.
Drill it out for some BIG bolts, maybe?

 

[r6155]

wrong design, fabrication and inspection.

1)Excessive dimension of fillet welds
2)The two external vertical ribs must be located at maximum distance from vessel vertical centerline.
3)See good design of saddles in PIP standard VEFV 1106

Regards
r6155

 

[KLee777]

Thanks for your responses, all. We have found four total vessels so far with gaps ranging between 1/2" to 5/8". All from the same manufacturer.

dhengr, even if I was comfortable leaving the fixed side of the drum alone, I cannot shim the floating support, which also looks like this. I won't risk it binding up. Also, I'm less concerned with the stresses in the saddle than I am about the stresses in the vessel itself due to uneven weight distribution.

jtseng123, yes, this is obviously from weld cooling and subsequent shrinkage. This could have easily been avoided and the OEM admitted they are in the wrong. In comparison, the way our local fabricators keep this from happening is to bolt the saddle being welded to a rigid element before making the shell-to-support welds.

I have walked these out in the field this morning with a local fabricator I trust and a couple of my colleagues. We are going to proceed with repairing these by jacking up one side of the vessel at a time, cutting off the existing baseplate, trimming everything back to level, and adding a thicker baseplate to make up for the cuts so we don't change the elevation. This can all be done in situ, which is good news for the project schedule. Additionally, we won't be touching the welds directly attached to the vessel.

The OEM has agreed to the backcharges to repair these, at least. They know these should never have left their shop.

Thanks again for your input!

 

[tankSnvessels]

Sounds as a very practical approach. Remove the base plate (lower most) and add gussets from vertical supports if required.