Belleville washers effectiveness on 6" to 60" flanges (pipe and HX) myth or fact 1

[ElroyJetson]

Hello everyone,

I was wondering if belleville washers really work effectively.

1. All springs add or maintain force but how much compared to relaxation of bolts and gasket?
2. How is the variance of a torque wrench factored in on bolting them up(Given values of torque manufacture is different than actual)? In other words how much can you be off by and it still be effective?

Please respond with technical data, studies or calculations. My experience is they usually seem to work only because the flange is inspected more carefully, torque wrench is used, Flange make-up is more closely watched higher torque value is used or a new or better gasket is put in.

 

[BigInch]

I don't see where they make anything for piping. Nothing about making them to ASTM A 194
Like these for example,

http://www.portlandbolt.com/technical/specifications/astm-a194/

 

[racookpe1978]

Flanges should be made up per the gasket manufacturer's specific torques and procedures.
Anything past that exceeds prices and does not make the gasket any more reliable.
Anything less makes the gasket less reliable (almost certain to leak over time) regardless of anything else done.

A metal-to-metal flat-face connection relying only on the force across each stud - as on a turbine head assembly or generator casing enclosing hydrogen gas? Follow the manufacturer's specific written procedure to every decimal place. Or it WILL leak. And you WILL pay to rebuild it to the turbine maker's specifications.

 

[ElroyJetson]

I am not in the Aerospace industry but rather in the Oil and gas or petroleum. I don't no how to change the tag on my user name.

To be more specific I am talking about putting Belleville washers on the bolts of the body flanges on a ASME Section 8 Div 1 pressure vessel (Large piping rarely used on). I am not concerned with how to stop a leak but rather taking them off because they may or may not be needed to maintain a seal. I am familiar with ASME PCC-1, WRC 538, and WRC 510.

Here is the link if your looking to buy them

http://www.ghxinc.com/ProductsServices/Fasteners/Belleville.htm

I have read studies written by the companies selling Bellevilles but wonder if there is any third party studies to confirm or deny.

 

[racookpe1978]

Belleville washers are "bellowed" washers that attempt to clamp permanently the stud by requiring an additional force be used on the two nuts (in addition to that force that is compressing the gasket) to squeeze the bellows flat. That additional squeezing force is expected to prevent the two nuts from backing off of their intended clamping force on the two sides of the flange. Lockwires, double-nuts, and pinging the stud threads are also sometimes attempted. They don't always work.

Stretching the stud, then spinning on the nuts to the required position, then releasing the stretching force so the nut prevents further release of the stretched stud is the best way to get an accurate clamping force on the PV flange. Then, you "only" need to maintain that clamping force.

In your application, it is far better to use a positive clamping (anti-loosening) process such as Nordlock washers . An advertisement is from one typical sales page below.

Nord-Lock Washers

The NordLock® bolt securing system uses geometry to safely lock bolted joints in the most critical applications. This unique wedge-locking system offers the following advantages: - Maximum safety - Resists loosening caused by vibration and dynamic loads - Ease of assembly and disassembly - Locking function is not affected by lubrication - Positive locking at low and high preload levels - Controllable preload - Same temperature characteristics as standard bolt/nut - Reusable Nord-Lock® washers with enlarged outer diameter (sp) in combination with flanged nuts/ bolts are recommended for use on large/long holes, painted surfaces or soft materials, e.g. aluminium. The key is the difference in angles. Since the cam angle " a " is larger than the thread pitch " ß ", the pair of washers expand more than the corresponding pitch of the thread. NordLock® washers positively lock the fastener in a joint which is subjected to extreme vibration or dynamic loads.

They are expensive if used on every pipe flange, but reasonable i fyou are putting only on PV vessel flanges. Anti-loosening compounds are also available (both red (permanent) and blue (releaseable) locking chemicals are much less. Are you worried about chemicals degrading at the process temperature?

 

[BigInch]

No doubt way too expensive for me.

 

[desertfox]

Hi ElroyJetson

Here is a link to an article which shows example calculations of Bellville washers in a flanged joint and a busbar joint.

http://www.sealing.com/fileadmin/docs/Using_Bellville_springs_to_maintain_bolt_preload.pdf

 

[hacksaw]

the benefits or disadvantages of various washer designs are covered in detail in the PVT, quite detailed experimental studies.

 

[11echo]

I work for a company that put "Belleville washers" on all their main steam headers. They were having problem with connection points loosening up and the steam would leak and cut the flange face. Main problem was they weren't follow their own program to go back the re-torque these connection points on the scheduled times. They did this for a few years, bottom-line these washers did not reduce any leakage, and now added a significant cost the projects ($7.50 each 6 years ago, for the sizes we were dealing with), so they dropped the requirement.

 

[3DDave]

Thermal cycling is the main problem to be overcome with gasket seating a close second. The flange will heat faster than the bolts; if there is too much difference the additional stress in the bolt or the gasket will cause deformation. This lowers the pre-load and allows a leak. Unless the bellevilles can produce the required preload without bottoming out, they are useless.

I can't see where Nordlocks would make a difference either. If loosening is only a matter of turns retention, they would help, but they won't help with tension retention, which is the goal in fastened, gasketed joints.

From a plain fastened, thermally cycled, pressurized joint design - I'd look at the gasket characteristics and determine hoe to make the fasteners about 1/10th the stiffness of the gasket to give some compliance, particularly for thermal transients.

 

[sjrfc2]

I've seen one company advertise a line of Belleville's that were designed for pipe flanges that, when taken to flat load, would result in a specific stress in the bolt: 45 ksi or 60 ksi etc... I think it was Key Bellevilles. I thought it was a clever way of using simple tools to provide feedback that the bolts were tightened relatively evenly and at the right load. When I was looking into use of Bellevilles, it was for a pressure-assisted gasketed joint (boiler handholes on a superheater). After going to operating temperature and cyclic pressure loading the spiral wound gaskets would creep and the joints would loosen. We tried Bellevilles there. They seemed to work OK in that application but we never pursued them for pipe flange joints. Although spiral wound gaskets can creep, installers should have a good feel for bolt load by verifying the raised face is coming into contact with the gage ring.

Some valve manufacturer's seem to make them standard items for valve stem packing gland bolts as well as pressure seal bonnet retainer bolts (especially with graphitic pressure seal rings).

If you visit the Rocket Garden at Kennedy Space Center in Florida you'll see them on rocket engines.

 

[Gator]

The mention of the rocket engines reminded me of seeing Bellevilles at the warm/cryo piping interface of main exchanger air separation unit cold boxes.

Piping Design Central

 

[EdStainless]

I used them on flanges that saw both temperature and pressure cycling.
We used one on each side (studs) of the flange.
They were slightly different ratings. We torqued to just flatten the weaker of the two, leaving the stiffer one to allow for expansion.
They are a pain. They are expensive and you must use good hardened steel washers under them and make sure that they are free to scoot (add a bit of lubripate).
We had much better sealing when we used them, but I am not convinced that conscientious installation and periodic re-torquing wouldn't have done the same.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[gasketguru]

Some people touted disc springs as a universal panacea for a while, but they often add more complication than do any good. The springs used on bolted joints tend to be quite thick with a low dish height, and if tightened flat have rather a high internal stress. You also need to get ones in a material suitable for the application temperature (e.g. so they don't lose spring temper) and these can be costly. As has been said above, thermal cycling is what kills a lot of bolted joints - all sorts of tricks have been used over the years - longer bolts and extension sleeves as well as waisted bolts to adjust overall joint stiffness - you need to analyse what is happening overall in each particular case.

I'd suggest you get hold of a report from a few years ago from:- ASME Standards Technology LLC, Called "Bolted Flanged Connections in Elevated Temperature Service" reference "STP-PT-036" which is interesting, but also concludes that using Belleville washers isn't much use on high temperature creep problems.

 

[TensionvsTorque]

In my experience i would say belleville washers are more smoke than substance. Any flexible load considerations are nominal, however I always see them in spec and iso drawings.

 

[oldrunner]

We are considering using Belleview washers on transformer bushing flanges. The flange has a non-constrained gasket between the bottom of the flange and the top of the transformer. The reason is to compensate for the compression and relaxation cycles during a seismic event. Any comments?

 

[C2it]

I don't see the value of bellville washers in flange bolting at all. The bolt itself is usually sufficiently elastic under tension loads. If there is a question due for instnace to cycling, then add a spacer, as its commonly done in pressure vessel flanges, to increase the tensile bolt length under load.

 

[israelkk]

The only way I see Belleville spring's advantage is to use Belleville springs designed such that they have a deflection range where the spring force is constant. This happens when the spring proportions are h/t=1.41 (h = deflection to flat, t = spring thickness). Such spring can be loaded to maximum deflection where the load is still constant. Thereby, allow stretch of the bolt due to relaxation, temperature changes even for millimeters or tenth of millimeters depends on the load and the space between the bolts.