belleville (disc) spring 2

[dho]

any one knows the production process ?
we buy it from one maker. during the years, the i.d. gets bigger and bigger. we match our part to the disc. now suddenly they change the mold and the i.d. is .010" smaller even it meets their spec in either case as they call the i.d. as .XXX min. their quality is not as good as we like particularly the burrs on the i.d.
we asked another vendor to custom made for us. unfortunately, it takes a set when pressed to flat.
so, I like to know is there any special process to make a good Belleville (disc) spring. or just our new source need to find a much stronger material, I mean special rolled 302.

material 302
old source, HRC45. the disc spring height takes a set on less than .001 when pressed to flat. (free height .039)
new source, full hard, 180K tensile. HRC41, set .004", of course, the load diminishes.

since 302 is not heat treat-able. does it tell me must find 302 sheet with 215k tensile?
thanks.

od, 500, id, 250, t .025, free height .039.

 

[tbuelna]

Here is a good tech reference that should answer your questions.

You might also refer to DIN 2093 which is a spec covering quality of disc springs. Based on your OP, DIN 2093 Group 1 quality (stamped/cold formed/corners rounded) would be appropriate.

It is common with compression springs that operate with very little margin before solid height (or coil bind) to be compression set by the manufacturer. This provides more consistent performance from the spring in service. The stress level (215ksi) you describe seems a bit high for 302, so you might consider 17-7PH instead. The tech reference linked also provides some design tweaks you can use to improve the performance of your disc spring design.

Hope that helps.
Terry

 

[israelkk]

It is probably was made of 301 spring temper strip. 302 comes aS full hard 40-45RC. Therefore there is no guaranty that all batches will have 45RC. As tbuelna mentioned, good quality belleville springs should be compressed to flat few time to remove most set during manufacturing.

 

[WKTaylor]

dho... this is a bit rambling, hope the main points get across...

Part of the problem of buying a part number [PN] from a vendor is that You are at the mercy of the vendor: They may/may not change the dimensions, materials/heat treatment, cold working, manufacturing process, etc over time... simply because the can... it's their part. However, IF the part fails to meet the vendors' own published dimensions and/or materials specifications, then shame on them for selling You an out of tolerance part. Suggest reviewing the vendors tech data sheet...the one Your purchased the part-against X-years ago and the currently available TDS on-line or in a brochure, etc... against the parts as-received. CAUTION: devil can be in the details [IE tolerances, options, materials, fab processes, etc], and there could be weasel words such 'as subject to change without notice', etc.

This applies for any vendor supplied commercial Item [Commercial off the Shelf - COTS or Commercial Item Description - CID] product: toilet paper, metals, plastics, adhesives, rivets, paint, cleaning compounds, fabric, etc, etc, etc. IF your item was made against a known spec then the vendor is obligated to have the item fully in compliance with that spec; otherwise it’s the wild-west.


Unless an 'item' is purchased against a known rigid specification, that requires certification testing, such as a federal, military, ASTM, SAE, etc spec, then variations are known to creep-in. Also, published specifications for vendor products are notoriously inaccurate about details; and may/may-not have any certification/quality verification testing done on-them. In many cases the TDS for a customer is bare bones and does not include essential manufacturing/conformity details.

NOTE.
Be extremely cautious if a vendor states that ‘PNxxxx has been inspected or tested [etc] per spec xxxxx’: This statement doesn’t mean it has been inspected, tested and certified to meet spec xxxxx… only that the vendor has picked/chosen elements of a spec to test their product against… and can subsequently use that spec tangentially [illegitimately] for their TDS/advertisement purposes.

Hmmmm... this situation is definitely NOT what we want in aerospace: unpredictable configuration and quality are positively unacceptable. It is essential for design/analysis/testing to validate a part ['item'] is suitable for use on an aircraft; then that item be purchased, installed and operated within known [not unknown limits]. To be of any use to us in aerospace a COTS item to be 100% predicable/testable/repeatable/reliable.

WARNING
Sub-tier vendors… especially those in the PRC… are a known to be extremely careless in their specification controls and fabrication documentation. I have seen this first hand: that is why current USA federal regulations state that anything installed-on, or used-in, an air vehicle, must have manufacturing traceability and certifications.

IF this item is not used in an aircraft or any flight related item... such as support equipment, tools or manufacturing... and strict controls are deemed important, then suggest You specify an item out of the catalog or TDS by part number; then repeat those critical elements [in this case dimensions, surface finishes, materials, spring-rate, etc] on Your purchase contract; or have them re-number Your item with a special identifier number to ensure the vendor is absolutely committed to produce that part; or procure that same spring from the vendor with Your drawing [P/N] of their spring with added details to ensure consistent manufacturing/quality… and indicate that ‘P/N xxxx’ may be purchased if it fully conforms to this drawing [company certification], then that PN would be acceptable’. OH yeah… and strongly suggest You mandate the vendor certify the shipped item as conforming to the TDS and contract or drawings, etc. IF a vendor will not comply with conformity documentations, then ‘Run-Forrest-Run”...

NOTE.
Spec-example of a Belleville spring…
MIL-W-12133/2 Washer, Spring-tension, Belleville Spring, Corrosion Resistant Steel

NOTE.
SAE design/usage manual has a lot of useful design, manufacturing, application info.
SAE HS 1582 Manual on Design and Manufacture of Coned Disk Springs (Belleville Springs) and Spring Washers



Regards, Wil Taylor

o Trust - But Verify!
o We believe to be true what we prefer to be true.
o For those who believe, no proof is required; for those who cannot believe, no proof is possible.
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion"]
o Learn the rules like a pro, so you can break them like an artist. [Picasso]

 

[tbuelna]

wktaylor- Great comments.

I have worked on a couple projects at startups where the people involved had minimal understanding of aerospace engineering practices. For any component like the disc spring mentioned, I always make a vendor item control drawing that includes acceptance requirements for materials, dimensions, performance, etc. The vendor item control drawing allows the component to be purchased using the vendor p/n, but it also requires the vendor to supply a C of C verifying the component still conforms to drawing requirements. This prevents the vendor from supplying components having their existing p/n, but that have been changed in some way that would not be acceptable for the intended application. Most reputable vendors are quite happy to supply a C of C with each order for a couple bucks extra.

The problem I typically encounter at companies with limited experience in aerospace is that management always wants to know why is it necessary to spend their precious engineering budget on vendor item control drawings when they could simply place a PO for the COTS part without the drawing? I try to explain the reason for doing so. Sometimes they understand and sometimes they don't.

 

[dho]

change material is a little bit late in the "game" as it will be class 1 change.
SCHNORR HDBK is a good tech ref. it also said pre-set.
but i still think the problem with our NEW source is the material "not" strong enough. its HRC is about 41 compare to the old source 45.
we tested several times on the old disc. confirmed HRC45.
the 215k was approx conversion from HRC45.
btw, MIl-S-5059 specifies full hard 302, 185k minimum.
thanks to all.

 

[israelkk]

dho

Can you give full spec (dimensions, forces, deflection) of both springs?

Is the old spring a standard spring from a catalog? If yes who the maker?

The dimensions you posted "od, 500, id, 250, t .025, free height .039." seems odd. Do you mean OD=0.500" and ID=0.025"?

The height 0.039" to thickness 0.025" ratio (0.039/0.025=1.56") is too high for common belleville springs, especially if you intend to use a stack of springs.

What are the deflection and forces you require from the spring.

 

[dho]

sorry, i did not notice missing period in the numbers.
MFR, "As..ted Spring", from catalog, material AMS5906 type 302 chemistry only. minimum I.D. 0.255, maximum O.D. 0.500, stock thickness 0.025, H approx, 0.038, H1, 0.031, load at H1, 85-105 lbs. P (calculate load at flat position) 160 lbs.
we bought this disc spring for thirty, forty years. the I.D. getting bigger and bigger. we adjusted our parts to meet the I.D. to keep clearance at 1.5%. now, suddenly, the I.D. goes back to .255 as the catalog allowed. all our parts not fit.
in addition, there are burrs on the I.D.
so we contract other to make one. we said to make exact as the one we have. since we hardness tested "AS" disc, and got HRC45. we say it should be made from full hard sheet. from my previous threads, you know the outcome. (they did nice job on de-burr.)
the disc spring used in our pressure switch application the load is adjusted. but you know, the spring should not take a set and creep during the time.

 

[israelkk]

dho

You may simply solve your issue. Take few springs and press them to flat and hold them for 24 hours. If the spring works in high temperatures put the pressed spring in an oven at that temperature. After 24 hours take of the load and measure the set that the spring/s had.

Now repeat the process and press the same springs for 24 hours again and measure the compression set. It should be zero or no more than 2% of the total deflection to flat.

If the results as I expect, take all your springs and press them to flat for 24 hours before installing them in the valve. You can easily do the press using a bar with threads on its both sides. Install all springs (in series) on the bar and use two nuts on the ends to press all the springs to flat springs together. If you want to be strict, place a thick flat spacer between each pair of springs. This way you will not need to precisely align the springs to each other.

 

[israelkk]

dho

The company you referred to is the best company and most knowledgeable I know off (at least they used to be) however, I had my share when they refused to do some processes we asked them. They probably think they know better than the client. I believe if you wanted consistency then you should order custom springs to your requirements (if they will agree to it at all) but it will cost you much more.