Plate stress - vacuum pressures 1

[nuche1973]

Greetings all, I have a particularly interesting problem. For several months I have been checking the design of dust collectors. This involved verifying the housing thickness with respect to the vacuum and deflagration pressure, for both round and rectangular collectors. As you may be aware, these structures are under 15psig and do not necessarily fall under ASME BPVC. However, I have been using ASME BPVC as a guideline, especially for the round collectors. My problem revolves around the rectangular collectors.
Up to now, I have been using simple beam theory per AISC for these collectors. Under AISC, I have to mathematically determine the effective width of the plate what the plate contributes to a plate with stiffeners combined section and determine the section modulus. I then, use that section modulus to determine the allowable pressures for both vacuum and internal pressure based on the spacing of the stiffeners. I use both the high and low section modulus, since they are not the same when the combined section is in negative or positive pressure.
I was recently informed that my results were extremely conservative. This came about when my client sent me a drawing asking for the maximum operating pressure. I determined the maximum operating pressure to be of -9 inches of water column. However, the collector is operating at -20 inches of water column. A coworker suggested to me that I refer to Roark’s Formulas for Stress and Strain, as a check. When I used the formulas outlined in Chapter 11.6 Thin Plates and Stiffeners, the resulting pressure was higher, a lot higher (around -300 inches of water column!)
The formula in Roark’s is similar to AISC, with respect that the effective width of the plate is required. However, the calculation gives a larger number than AISC (7” vs 3”). This is obviously where my discrepancy lies. However, I want to be sure that I am looking at this from the right perspective. I am curious to hear how a mechanical engineer would address this issue. Please bear in mind the following:
1. I am aware that I am a structural EI and I am not pursuing this type of work. Although I enjoy it, but these projects were pushed on me (work is work and lowest man syndrome).
2. The company I work for has a Mechanical division but work mostly in HVAC (I got blank stares or the usual “why are you doing this?”)
3. Most of my work is by hand, or using MS Excel. We have RISA but limited licenses.
4. And as always any comments are greatly appreciated.
Thanks.

 

[JStephen]

Going from -9" to -20" sounds reasonable. Going from that to -300" sounds like there's a mistake somewhere.

It sounds to me like the basic issue is that you are using a very approximate solution to a problem. If the solution is not codified somewhere, then other people are going to use different approximate solutions with different assumptions and get different answers.

The general problem could be treated as beam spans, as 2-way plates, as 2-way plates with large deflections, as a finite elment problem, as a lab test, etc. Generally, if you get an easy solution that allows economical construction, you use it. If the easy solution gives you expensive construction, then you start looking at it in more detail.

 

[DSB123]

Rectangular ducts with stiffeners are normally addressed as plates as per Roark Formulas for Stress and Strain. However if you are using the basic flat plate theory from Roark then you must remember that the stress in the plate may not be the limiting factor. For bending of flat plates, for the formulae to be applicable, then the maximum deflection must be less than half the thickness of the plate. If it is greater then membrane stresses predominate and you must use the formulae for large plate deflections from Roark.

 

[gr2vessels]

If you are using ASME VIII-1 for the circular vessels, why don't you use the ASME VIII Appendix 13 for the rectangular vessels, just for consistency.
Cheers,
gr2vessels

 

[nuche1973]

Thanks for the recommendations. I appreciate it.

 

[r6155]

Calculations is not the 100% of the work.

Think on Inspection and Testing Plan and Dimensional Tolerances

Regards

r6155

 

[Geof]

I once did design work for bulk storage silos. It's been some years since I've done the calcs but...

Generally when a baghouse, dust collector, or any other item failed, a stiffener buckled. We generally limited bending stress in panel supports to avoid this. Other tactics, where possible, were to turn angles leg-in, use closed sections, etc.

A good reference was from the Lincoln Arc Welding Institute. I don't recall the title but check the site - the references are all very inexpensive.

 

[carnage3]

There is one very important thing to remember. Factor of safety in vacuum chambers can be much lower because if they fail they don't create shrapnel. I would expect that the pressure vessel calculatoins would already have a significant factor of safety built in. very likely more than you need. a large portion of your discrepancy could be that one set of calculatoins has the FOS built in an the other doesn't.

 

[jamesl]

Appendix 13 in ASME BPV code does not cover vacuum pressure in noncircular shape.

I think Geof refers to the book "Design of welded structures" by Omer Blodgett. That is a very good reference book.

As a mechanical engineer in pressure vessel design, we use this book and Roark's to do calculations on structural components.

 

[saplanti]

Your problem extends to pulsation issue as well. Dust collector pulsation frequency is a big factor to determine the thickness of the plate and the stiffeners.

So you must have a deflection restriction due to pulsation on the plate and stiffeners as well as pulsation frequency.

You'd better ask the pulsation frequency and get away from it at least around 20-25 %. You will see that this may increase the plate thickness depending on the stiffener distribution, and the stiffener sizes.

Plates and stiffeners shall comply with static loads such as self weight, insulation, and pressure (internal pressure or vacuum), and the pulsation criteria.

Hope it helps.

Ibrahim Demir