Squareness of a spring 1

[sbozy25]

The spring in question is an extremebly high stress, very high cycle requirement, top of the line material, bee hive shaped spring. What I mean by that is both ends are smaller than the middle section.

For the past 15 years we have been measurign the squareness of the end by setting the part on a flat surface and laying a digital protractor on the top and taking the reading. I realize this is not correct, but that is what we have done for years without complaint. Well, now our customer is arguing with us about our parts, they say they are out of square and we say they are in.

They measure the part by setting round cylinders in the end coils with a shaft between themselves. The shaft then centers it self and pushes on an idicator that reads a degree of square. (sorry no pictures, only saw the device once myself). My issue with this is that the ends of the spring are not true circles, the ends are a varying pitch that forms a Spiral of Archamedies.

I can not accept their method of measurement, and they will not accept ours. Anyone have any ideas how we might check for squareness?

 

[TStaples]

You did not indicate a size, but assuming the spring is small enough, I would recommend mounting the spring on a precision gage pin that correlates to the smallest I.D. of the spring. You can then locate that on an optical comparator (Simple V-Block), and measure the relative angle between the ground ends, and the theoretical centerline (the gage pin.) I would think the main issue is to not compress or deform the spring during measurement.

-Tony Staples

http://www.tscombustion.com

 

[CoryPad]

I can't speak for it personally, but have you considered Testing and Tolerancing from SMI?

http://www.smihq.org/public/publications/encyclopedia.html#Anchor-Testin-15664

The price is right at $9.50, and it says that it covers squareness and parallelism.

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[sbozy25]

I have all those books. The president of SMI is my CEO, so we have it all.

The problem is that the ends are an ever expanding spiral, so there is no good way to find a centerpoint to test off of. This is a sort of un-conventional spring in a manner of speaking.

 

[sreid]

Is there some way you could use a CMM to probe the ends of the springs. And then reach some agreement with the customer about what the planes of flatness are?

What I'm really suggesting is to find a third way to make the measurement where you and the customer might reach common ground.

The hardball way to handle this (and bring it to resolution quickly) is to agree to test the springs with their test method and let them know what the price increase is going to be based on the new reject rate.

 

[desertfox]

Hi sbozy25

What tolerance on squareness are you working to?
We may be able to suggest better mearsuring technique's once we know the accuracy expected.
According to my associated spring design handbook a commercial tolerance on squareness is 3 degrees.
It also shows a method of measuring squareness by standing a spring on end on a horizontal flat surface and bringing it upto a straight edge vertically positioned off the horizontal flat surface, the spring is slowly rotated and the gap between the body of the spring and vertical straight edge is measured and recored as the spring is turned, the spring is then placed on its other end and the same procedure is repeated.
They also check parallelism by standing the spring on one end and recording the variation of the free length on top of the other end by taking measurements around the circumference of the top coil.
In your case with a barrel shaped spring you would need to bring the largest diameter coil in contact with the vertical
straight edge then rotate it and measure any gaps as they occur.

regards

desertfox

 

[sbozy25]

I am working with 4.5° which is the standard for "power" springs. So I have from 0 to 4.5 to be within spec. The spring measures around 9-12 inches in free length and has wire size between .5" to 1" with an OD around 4 - 5". They are rather large springs.

I actually have to laugh at your suggestion of how Associated spring guide tells you how to measure. We tried that, and it didn't work. Furthermore I have to laugh, because my boss, actually wrote that when he worked for Associated years ago. Between him and a few other, they wrote that design manual.

I think I agree with sreid, and use a CMM to generate the planes. I will just need to make sure the inspector understands that the part does not have true circular ends, so they need to adjust for the offset....

 

[israelkk]

The squareness of the spring ends are with reference to the spring axis. At least this is how it is defined in the MS standards for helical compression springs (Military and aerospace springs). Therefore, the spring axis should be defined and established first before the squareness can be measured and verified. As I see it, the spring should be placed on an accurate guiding shaft protruding from both sides of the spring. Place a plates with a hole larger than the shaft diameter on one end of the spring and measure the angle between the plate and the shaft. Repeat the procedure for the second end of the spring.

 

[TVP]

Apparently everybody used to work at Associated Spring at some point...Associated Spring still uses this method for checking squareness, with a purpose-built gage. What part of this method didn't work? I agree with you that using a CMM will probably be the easiest way to agree upon a "referee" method.

 

[sbozy25]

israelkk, thank you for that idea, I will give it a try and see what I can come up with.

Yeah, it seems like everyone worked for Associated at one time. From what I hear, they were one of the best companies to work for back in the 70's....

Everything in that method didn't work, the way the springs arr manufactured, they have a slight lean and it skews that method of checking.

 

[israelkk]

sbozy25

I didn't work for Associated springs (Barnes) but have great respect for their role and contribution to the spring design and analysis area. When the plate sits flat on the spring end it defines the end flatness (three points in space define a flat surface). Any other method like CMM prove to be false. I fail to see how it will determine the exact three points on the spring end that will touch the flat plate (however I may be mistaken).

 

[israelkk]

Correction:

"Any other method like CMM "MAY" prove to be false."

 

[btrueblood]

Agree with sreid, "The hardball way to handle this (and bring it to resolution quickly) is to agree to test the springs with their test method and let them know what the price increase is going to be based on the new reject rate."

Your customer has a requirement for the geometry of the spring, in order for it to suit their end use. They've told you (or can tell you) what that requirement is, in the form of a gaging method. Regardless of terminology (squareness of a doubly-tapered helix?!), they want it their way, so cost it out and send them the updated pricing. They can decide whether to pay you or let you continue doing it your way.

 

[sbozy25]

hahhah update pricing.... that is a good joke!

I'm afraid if I were to reccomend that, I would be seeking employement in the very near future... Others in the company have warned me about that on this particular product line....

I have a meeting with them in December, at that time I am going to poke all sorts of holes in their gauge, and get them to see the light... (hopefully)...