25 lbs preload spring

[bernardg]

Hello all,

We make dc motors. Our customer's requirement is that we pass a 10 G vibration test at 24-1000 Hz frequency ~ the motors being excited along the axis of the motors for 6 hours.

We manufacture shorter stack motors whose armature mass is 1.3 lbs and the larger stack motors whose armature mass is 2.2 lbs. (The 1.3 lbs armature @ 10G vibration would exert an axial load of 13 lbs-force.... and the 2.2 lbs armature @ 10G vibration would exert an axial load of 22 lbs-force.)

So we use a 25 lbs preload spring (25 lbs @ 0.062 working height) on the shaft end to take up any tolerances and keep the armature in place without impounding on the endbells. Our design allows for a 0.062" working height for that spring.... and we run consitantly around 0.0585" on the lower stack motors and 0.0490" on the larger stack motor.

The preload spring broke on both the motors during the vibration test... and we donot know why. I could only assume that there are some other factors that would aid in this other than just the armature mass.

Kindly go through the calculation that we did to come up with that 25 lbs preload spring...

Load = 25lbs
Working height of the spring = 0.062"
Spring rate = 620 lbs/in

freeplay for the preload washer on the 1.3 lbs mass armature = 0.0585"
25 + (0.0620 - 0.0585)*620 = 27.17 lbs-force

freeplay for the preload washer on the 2.2 lbs mass armature = 0.0490"
25 + (0.0620 - 0.0490)*620 = 33.06 lbs-force

The thing that confuses me is that...
The 1.3 lbs armature @ 10G vibration would exert an axial load of 13 lbs-force on the preload spring.... and the 2.2 lbs armature @ 10G vibration would exert an axial load of 22 lbs-force on the preload spring. Since the preload spring would exert a 27 lbs-force on the shorter stack armature and 33 lbs-force on the larger stack, there should be no armature movement.... so there should be no compression/expansion of the spring... so the spring shouldn't fail. BUT WHY DOES THE SPRING FAIL? IS THERE ANY OTHER FACTOR THAT WE SHOULD TAKE INTO CONSIDERATION? KINDLY ADVICE.

Thank you in advance for all your suggestions.

 

[pjhype]

mpoc,

Could it be that during vibration testing you are exciting the natural frequency of the armature on the preload spring? According to the numbers you are providing, the 1.3lb armature will vibrate at 68 Hz and the 2.2lb armature will vibrate at 53 Hz. If you can provide the value of the vibration PSD at these frequencies, you can determine the amplification factor for the load acting on the spring.

pj

 

[ccw]

Hi Mopoc,
You stated:

"...so there should be no compression/expansion of the spring... so the spring shouldn't fail. BUT WHY DOES THE SPRING FAIL? IS THERE ANY OTHER FACTOR THAT WE SHOULD TAKE INTO CONSIDERATION? KINDLY ADVICE."

As we hoped to show in thread: 404-67465, the 1.3 lb armature may exert MUCH MORE than 13 lbs-force (1.3 lbs x 10g) due to mildly damped or undamped structural resonance in the motor structural system being excited by the sweep. I am not sure the subject motor configuration here is the same one in the earlier thread (where you said the preload spring was a wavy washer), but the explanation of dynamic forces would be the same.

I would take a WAG that, unless the configuration is changed, or the acceptance criteria is relaxed, the preload springs of these subject motors will continue to fail unless it can survive 40 X armature weight applications for 100,000 cycles. On the other hand, how could you prove me wrong? (Hint: SDRC, MTS, Brown-Alamang @ UCinn, etc.)

 

[desertfox]

hi mpoc

can you provide dimensions of the spring, i maybe able to help, i need the following information:-

free length of spring

outside dia of spring

material of spring

number of total turns

end condition of spring ie :- ends of spring closed
and ground


regards desertfox

 

[RHD]

You do not mention the profile of the vibration. For example if you are using a random on random profile, the peak G's can be 4 times the RMS value.
If you are just vibrating in an open loop mode, ie no control accelerometer, then I would look for a resonate condition. You can also examine the shaft to see if there ia any fretting corrosion. This rusty looking debris will indicate some relative movement between the brg and the shaft.

 

[bernardg]

Hello pjhype,

I am not following how you found out that, the 1.3lb armature would vibrate at 68 Hz and the 2.2lb armature would vibrate at 53 Hz. Kindly explain it for me plz.

With respect to the PSD at which the vibration table is excited...

Frequency PSD
24 0.06
70 0.50
100 0.50
250 0.10
1000 0.025

G (rms) = 9.86

Kindly help me to find the amplification factor.

Thanks,
bernie

 

[bernardg]

Hello Desertfox (Rommel )

To answer your questions..

free length of spring = 0.100"
outside dia of spring = 0.84 +/- 0.02
material of spring = carbon spring temper steel
number of total turns = 1
number of waves = 3

working height = 0.062"
Load at 0.062 = 25lbs
spring rate = 620 lbs/in.

Kindly let me know if you need any more information regarding the spring. I can email the drawing of the spring if you want me to. My email id is bgruban@yahoo.com

Thanks in advance for all your help!

bernie

 

[desertfox]

Hi mpoc

I didn't realise at first it was a wavy washer you are using
for a spring, anyway all I need now is the material thickness of the washer and its inside diameter.

regards

desertfox

 

[pjhype]

mpoc,

f = [1/(2*pi)]*sqrt(k/m)

where k = 620 lb/in
m = (1.3 / 386.4) sl

This yields a result of 68 Hz for the 1.3 lb mass. This is valid only if you can model the armiture as being supported on the preload spring as a single degree-of-freedom oscillator.

pj

 

[Mendit]

Try using a spring with better damping qualities try this company.

http://www.smalley.com/spring_detail.asp

 

[bernardg]

Hello pjhype,

Kindly give me the formula to calculate the amplification factor.

This is the PSD data, our vibration testing center gave me.
Frequency PSD
24 0.06
70 0.50
100 0.50
250 0.10
1000 0.025

G (rms) = 9.86

Thanks in advance for all your help.

bernie

 

[pjhype]

mpoc,

I cannot tell you the amplification factor without seeing some test data. I would recommend that you perform a test where you would put an accelerometer on the armature and subject the motor to a sine sweep in the frequency range of interest. This will give you the transmissibility of the armature, and from that, you can determine the amplification factor.

pj