Test Plate Design

  • #1
I need help in designing a test plate for a leak rate test on a butterfly valve. The test must be performed in the field which complicates things.

The engineer that is helping with this has come up with a plate that is 25" in diameter and
1 1/2" thick carbon steel. This is not rational because this plate must be handled by persons and not any machine (ie. location doesn't allow any mechanical assistance, not enough room).
I have suggested an aluminum plate design. I need to know how to determine the size requirements?
Specifications: 20" butterfly valve with 20 bolt circle 1 1/8" bolts grade 8; must have 25" outside diameter and be able to withstand a 55psi force on this surface area.
The engineer that is helping is using a 75 pound flange rating or 150 pound ANSI. I understand the need to go by code, but is there a way for the engineer to design a flange to mount on the valve and use a lighter weight material? Actually using his brain and not some code. I have looked at some aluminum alloys (specifically 7068 and 7075) with an ultimate tensile strength of 83 to 103 ksi and yield strength of 73 to 99 ksi shear strength of 48 to 53 ksi density of .101-.103 lbs/cubic inch
The actual plate design is not the issue just the thickness and weight of the tool.
Any suggestions Greatly appreciated.
 
Last edited:

Answers and Replies

  • #2
549
28
A few thoughts.
By going with the code the engineer is "safe" if anything fails he won't be being asked very awkward questions.
The plate (blank?) that he proposes might well be an off the shelf item that will reduce lead time and cost, compared to a custom made model
How heavy would you aluminium plate be if it is still heavier then something that can be safely handballed the there is no advantage over a steel plate.
Go and talk to a couple of old hand mechanics and riggers on the site and ask them about moving the blank into position they'll probably say "no problem".
 
  • #3
Thanks for the reply jobrag. Yeah I know he is playing "SAFE" no problem with that but sometimes common sense and rationality has to play a role. The leak rate test has been performed by some "old hands" for years now with a single 3/8" carbon steel plate and about 8 large C-clamps (no rating known for the C-clamps). I have witnessed this test on a few occasions and questioned the safety from the get-go. Once pressurized the plate would force the C-clamps to begin to buckle and the Valves oring would blow out and the test would have to be started all over. Issues being clamps falling off, oring blowing out, and damage to opposite side valve oring surface, etc. This is when I began to suggest the two plate design using the valves existing bolts to secure the plates (aka flanges). Rigging the 100 plus pound flanges would be a nightmare adding countless hours to a shortened outage. There are multiple valves that have to be tested during our outages. Same sob story cutting back on personel and wanting shorter down time (shorter outages).
One other comment not once has there been a problem with the plate only the C-clamps. When this plate was fabricated no regard was paid to the safe design only the quickness of the test ie. get the job done. New rules now.
 
  • #4
Q_Goest
Science Advisor
Homework Helper
Gold Member
2,974
39
Hi bulldog fan. Welcome to the board. To answer your question, yes there’s a way to do it by the rules.

First, I’m assuming you’re bolting up to a 20”, 150 pound valve so the flange has to have those dimensions. Further, you’re looking for a blind flange and you want to rate the pressure at 55 psig. To rate it at that pressure you have to show that there is no way for the pressure to increase above that, either by providing a relief valve, set at 55 psig and capable of relieving the full flow of your source pressure. Or by showing that the source pressure can never exceed 55 psig. That, in a nutshell, is what you need to prove to show that the flange can be rated at 55 psig.

Second, if you’re in the US, and I’m assuming you are because you’re referring to everything in inches and pounds, then ASME B31.3 piping code will almost certainly apply to your situation. Even if it’s one of the other piping codes, they all generally have the same reference for pressure rating. Per B31.3, requirements for blind flanges is under para. 304.5.2. The piping code essentially directs you to Section VIII, Div 1, UG-34 with some minor changes in nomenclature. You can use the materials in B31.3 (you don’t need SA rated materials) however, so that makes it easier yet. For the sake of argument, let’s use 6061-T651 per ASTM B 209 as listed in B31.3, Table A-1. Stress allowable for that material is 14 ksi. That should be a relatively common material so no big deal obtaining it.

Next, you go to UG-34 and perform the calculation there. Factor C is per fig. UG-34 which would be 0.30. Joint efficiency is 1 since the plate isn’t going to be welded. And I’ll assume your O-ring diameter is not greater than 23”. Reducing this diameter to 21.8” or less will help because then you can use a 0.75” thick plate. If it’s 23” you need .79” thick plate so you’re back up to 7/8” thick. This type of flange is also called a “plate flange” since you’re just cutting it out of a solid chunk of plate. Anyway, this gets you down to a flange weighing about 50 pounds.

Because it’s being used on a piping system, it doesn’t need to get coded or anything like that. Just document your calculations and file them for future reference, so start by getting copies of the code and going through the paragraphs listed above and doing the calculations for yourself.
 
  • #5
Thank You, Thank You, Thank You O Goest this will help my argument. I'm sure you can tell I'm not an engineer so I have trouble where to look for the correct information that I need. Thanks Again.

Yes the test stand that we use has a relief valve so the 55psi is not a problem. I have been dealing with a young engineer who is already putting his application other places because he feels this job is to hard. He is the system engineer for the valves to be tested and was assisgned this task to come up with a test plate design. He comes to me and tells me to tell him what I need plate size, diameter, thickness, etc. He says he needs all this ASAP so his boss will get off his back. Has been an issue I identified 2 years ago and he has just now felt the urgency to fix it and our outage is in 4 weeks. Sorry to ramble on with my sob story I guess this is life; someone else doesn't want to do it, it needs to be done, oh I guess I can do it.
 
Last edited:

Related Threads on Test Plate Design

  • Last Post
Replies
4
Views
1K
  • Last Post
Replies
1
Views
2K
Replies
10
Views
1K
Replies
5
Views
6K
  • Last Post
Replies
5
Views
3K
  • Last Post
Replies
4
Views
1K
  • Last Post
Replies
3
Views
3K
  • Last Post
Replies
10
Views
7K
  • Last Post
Replies
2
Views
1K
  • Last Post
Replies
2
Views
3K
Top