max pressure seal design

[sherwin1]

How do you determine the max pressure a mechanical seal is designed for? I have looked on other seal company websites and next to their seal it states the max operating conditions the seal can handle. Below is a hypothetical example of what you might see on the website of a seal company.

Ex. Seal xyz is a cartridge single seal
operating limits:
pressure 0-350PSI
temp -40-450F
shaft speed 3600rpm
shaft size 1-5.5inch

Thanks.

 

[1gibson]

"A" seal is not going to handle 1" to 5.5" shafts. Or -40 to 450*F.

What are you trying to do? Where are you in the food chain? End user, EPC, equipment vendor trying to incorporate a seal, seal vendor?

To answer the question, I look at a pump curve, get the operating conditions, and send the info to the seal vendor for a quote.

 

[sherwin1]

OK i pulled a real word example and attached it to this post. I want to know how they came up with the max pressure the seal can withstand.

Recent graduate of Mechanical Engineering from a 4 yr university. Only engineer at the seal firm I work at.

I would think the numbers supplied in the attachment must be an estimate based on water because different fluids being pumped have different specs. But would like to know how they calculated it. I looked over many of case studies and have yet to find this information(equations). We don't have any FEA software and the CAD is from 1996 or before being run by a non engineer (self taught person). I would like to not worry at night when seals are sold that I worked on.

Thanks

 

[keyepitts]

You are in a difficult position. How did they determine MAOP of their seals before they hired you? There are too many variables for anyone to tell you how to evaluate a seal design on an internet forum, but with what you have to work with, I would recommend actual hydrostatic testing of any seals to at least twice the MAOP that you intend to list for the seal. Note that the hydro test does not necessarily guarantee that the seal will work under dynamic operating conditions. If you can test the seal dynamically, that should also be done, at least up to the MAOP, preferably 50% more. Materials must be compatible with the pumped product also, of course. As for the actual designing of the seal, if you have to ask....

 

[sherwin1]

They go to the competitors website and find a similar seal and call it good but had a previous engineer a long time ago(he took his stuff leaving no copies for me to go by when he left the company). In my opinion blindfold guessing which scares the hell out of me. I have purchased a goulds 3196 1 3/4 in shaft 1 inch discharge 2 inch suction, a 3600 rpm motor 220volt 3 phase motor, a frequency drive capable of going from 0 to 3600 rpm, a 525 gallon tank, and associated piping. I did this in an effort to have the ability to test these seals moving forward to cover my back side (CYOA). The test rig will be set up by the end of the week. Is there a specific procedure you would use to test the seals?

A little more about my background.
*As I stated earlier I recently graduate of Mechanical Engineering from a 4 yr university with a 3.5
*before college:
***14 years rotating equipment specialist for the oil and gas industry as a machinist/millwright
------equipment worked on:
=========pumps
=========turbines
=========compressors
=========blowers
=========gearboxes
=========etc.
***4 years USAF

Reason for the degree I was working 16 hour 7 days a week on a compressor crew and wanted to see more of my son and wife.

 

[sherwin1]

Also, is there any software that can help me get the information I am looking for and also help with the design process?

 

[keyepitts]

Try API682 standard for starters, googling for that or "API 682 Testing set-up", or similar should provide some inspiration.

 

[sherwin1]

I have been using API 682 thus far but was hoping to be able to calculate the specs I was looking for, for the seals then test them on the test rig as a proof not the other way around.

 

[keyepitts]

Digging around Google with the API682 term should turn up a few technical papers with seal design information in them, or pick up a couple of the mechanical seal design books available on Amazon. Still, for determining a MAOP with confidence, you will need to empirically evaluate your designs. Without FEA ability it will be very difficult to optimize the design, except thru expensive and time consuming iteration and testing to failure.

Edit to add- Most seal failures in centrifugal pumps do not happen due to overpressure, most failures are caused by heat in my experience.

 

[sherwin1]

Do you know of a good seal FEA software(designed for seals not like Ansys workbench which I used in college and didn't care much for "not very user friendly") that I can present to the owner of the company? I'll see if he would be willing to purchase a license but with the extremely outdated CAD software I'm not sure if he will bite.

Thanks

 

[keyepitts]

No, I don't know of any FEA specifically tailored to seal design, doubt any exists. You might want to get a modern CAD program like Solidedge or Solidworks with an integrated FEA package; probably looking at $5-8K minimum. Depending upon the materials you are investigating, their built-in FEA package may or may not work for what you need, ask their salesman to demonstrate a typical example before you buy. Even Autodesks Fusion has FEA available, and its low cost or even free, but is still early in its development and may not be suitable for what you want. Really, for most mechanical pump seals,your MAOP is probably going to be limited more by heat generated by face friction than by pressure causing rupture failures of seal components, unless you are building seals for high pressure (1000psi+) applications.

 

[sherwin1]

The only reason I was looking was remove the guess and have a known to work with since our sales force keeps getting asked by the end user the question of max pressure the seal is designed for.

Thanks

 

[Chpal1984]

Designing mechanical seals is relatively niche. Most seal manufacturers will have they own custom software for designing seals. I'm not quite sure what you are trying to do?

The seal manufacturer will tell you the max operating pressure for their seals. Call them up if you can't find it on their website. I wouldn't recommend you try and calculate it yourself. Lets assume that you can master the theory first time, with no one with experience to help you. Without knowing the material properties, you will only be able to come up with an approximate guess. This guess will be less good than what the manufacturer can tell you.

 

[sherwin1]

Chpal1984 as I stated earlier I work for the seal company and I am the only engineer. I am also a recent college graduate BSME. They have no way of doing this thus far and I am trying to solve this problem so that it doesn't cause them issues with end users. I have access to all of the material properties and detailed drawings.

What I am trying to do is answer the questions posed to me from the sales team that is being asked by the end users engineers. The questions are as follows:

What is the max pressure your seal design can be operated at?
What are the max operating conditions of the seal I am looking to purchase?

Our sales people look to other seal manufacturers websites and give them that information. This in my opinion is a bad practice and not a safe way of conducting business. It also means we don't find the safety of our customers matters or their equipment. (This is my opinion)

 

[Chpal1984]

Oh dear. I would suggest you test to the worst case conditions. Worst case temperature, pressure and speed. I'm guessing if you are the only engineer you don't have a huge seal range

In parallel I would build an FE model, i googled this look alright (

http://www.atlantis-press.com/php/download_paper.php?id=16998)

then use you test results to verify your model

good luck

 

[awhicker84]

Do you need FEA?

You don't have test results from the companies previous life? You have machining specs / dwgs... you must have more than 'nothing', no?

I would think a good '2D' code that was coupled with test results would yield good results or is the seal industry in fact very FEA driven?

 

[Chpal1984]

@awhicker84 no one needs FEA, you could do it all with testing if you wanted/could afford it.

Also when I say FEA I wasn't suggesting going to spend £10k on an expensive 3D FEA software (ANSYS ABAQUS etc). FEA is a mathematical method that can be applied to 1D, 2D and 3D problems, and there are some good cheap/free ones out there. You'll notice the example I sent through only has a 2D model.