Fine thread bolts - do they back off 12

[RCMRon]

These are 5/8 " fine hread studs holding 70 lb fan blade liners on a 500HP fan rotating at 712 RPM. the liners are held in place with 20 1 1/2 inch studs. When properly torqued, will the nuts back off and or shouyld they be tack welded or double nutted?

Does anyone know of others that use stud mounted fan blade liners?

 

[patprimmer]

80" is an unusual temperature.

The loads applied to the fasteners in both shear and tension are somewhat vague to say the least.

The actual size and number of fastners is somewhat vague.

The positioning of the fastners relative to load is somewhat vague.

While calculating stress is not my field nor one of my skills, I doubt anyone can even begin when you consider the lack of required informaTION.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[RCMRon]

again, thanks for the assistance.
To provide a few answers to the questions, what damaged the liner was that fact that it tore off of the blade at 720 RPM.

Operating temperature is approx 250 DegC, Diameter is 80"

I showed the pictures of the stud to a stud manufacturer, his assessemnet was weld failure.

This failure happend 2 hours after the fan was put back in operation after having new liners installed.(Explains good condition of protruding end of studs)

I attached a picture of the break in the weld. What I would like ensure is that the nuts do not back off for the duration of the time it takes the liners to wear. Our practice now to remove the liners is to arc off the nuts/studs and pull the liner off.

Thanks
Ron

 

[desertfox]

Hi RCMRon

I realse I am coming in late to this post but I have just had a look at the pictures you have provided, it doesn't look to me as though the welds have failed but the studs themselves in a brittle fashion.
The weld round the studs in your earlier post seem to be intact and also in your last post, I can see a weld bead around the base of the stud for at least 2/3 of the diameter which appears intact, after which the picture goes dark.
My thoughts are firstly does the stud supplier know that the studs are operating at 250 degrees centigrade and what analysis from your side as been carried out the determine what preload the studs will need, to operate correctly in service?
So in a nut shell what torque figure are you using and how as that been determined.
Failure can occur if the nuts are not tightened enough but from your posts I cannot see what external loads the studs are expected to see in service, or what stud preload your generating on assembly.
So we need more information as others have said to try and establish whats causing the failure before we can suggest better solutions.

desertfox

 

[RCMRon]

Hello desert fox, thanks for the reply.

When I sent in the pictures to a stud manufacturer (Different one than the studs used) the consensus was that it was a weld failure. This is a stud weld done with an old TR-1600 nelson stud welder.
Our torqueing procedures are inadequate so I can not tell you what they were torqued at.
This fan normally operates at 200DegC but in this case, only operated for 2 hours before the fan blade liner let go.

I'll add another picture of another broken weld/stud

In terms of loads, each blade is 70lbs, 20 studs roatating at 720 RPM on an 80" diamaeter fan wheel.

Regards

Ron

 

[RCMRon]

Here is a picture of a failed stud end of the failure

Hope this helps

ROn

 

[RCMRon]

On an earlier reply, someone wanted to see a cross section of the weld interface, I have this picture of one of the welds that did not fail before the backing plate let go.

 

[desertfox]

Hi RCMRon

Well if your torque procedures are inadquate then obviously they need sorting out pretty quickly.
Now I don't know how accurate I am going to be, but if I take the 70lbs and use that as a mass in kg and convert the radius to metres I can estimate the centrifugal force by this formula:-

F=m*w^2*r = 31.8kg * 75.4^2 r/s * 1.016m = 183.681kN

if each stud takes an equal load then each one takes:-

183.681kN/20= 9.184kN tensile load

The big assumptions here are the 70lbs acts at a 40" radius
and that each stud takes an equal load.
Now in addition to the tensile load the studs also see a side load (shear) due to thermal growth in the length of the liner, which without any freedom to grow will try to shear the bolts, we could estimate the shear force but we would need the original length of liner, material and coeff of linear expansion.
Finally the material strength at 200 degrees centigrade of both the studs nuts and liners will be reduced when compared to that at room temperature which may also contribute to the failure.
Another important factor is how the fan heats up, even with materials that expand at the same rate, unless uniformly heated, then stresses can develop through differential thermal expansion.
So even if you have a weld failure then you need to explore why, is it a defect in the welding procedures? or is it the external forces causing the weld to fail.
Why not send the studs to a lab for analysis and see what they come up with.

desertfox

 

[RCMRon]

Hello desert fox

I am told that regardless of the stresses put on the studs, the break should always be in the stud material and not at the weld. From what I can see, and others, the break is clearly in the weld area. We don't weld ourselves, we buy these from a machine shop that cuts the liners and stud welds the studs. People here have been reporting studs comming off for many years but never enough I guess for an entire blade liner to come off. With this and pictures, i suspect the welds. I should also send out for analysis. As for our torqueing procedure, we were installing the nuts with a 3/4 " impact wrench and the washer under the bolt was covering a 1 1/8" hole and being cupped. I believe that the design is flawed. I am told that a fit for purpose washer can be sourced. Any ideas? I attached picture of cupped washer.

 

[hydtools]

Quit using the impact wrench for the final tightening. It is unreliable for controlled tightening. Use the impact wrench at a low setting to run up the nut and then finish tightening with a torque wrench.

The impact wrench may be over-tightening the fastener and/or initiaing weld failure.

Use a thicker washer.

Ted

 

[unclesyd]

I think a complete redesign is in order.

One very big problem is that you evidently have movement at onset. Where it starts remains a mystery to the forum for lack of information.

Is this the original design or up powered version?
C
an you post a print or good sketch of the assembly?

 

[Screwman]

The close-up pictures of the failed studs show what looks to be a very brittle fracture surface and the one shot shows what looks to be inclusions in the weld area. I wonder if the person welding has the weld parameters all out of whack and is creating a massive heat affected zone. It looks like you have a very small portion of ductile failure and a large amount of brittle failure.

Another possible cause:
Is there any type of chamicals in the atmosphere that this is operating in? Given the short delay between start-up and failure, the idea of environmentally caused Stress Corrosion Cracking comes to mind.

You need to get a metallurgist who knows failure analysis in to look at this. The weld stud mfg. is going to say that it's the welds (of course) and the welder is going to say it's the studs (of course).

 

[desertfox]

hi RCMRon

I agree if the stud welding is done properly then it is the shank of the stud that should fail and not the weld, however we don't know what the external loads are on the liner so we don't know if theres an overload condition on any of the studs.
I found this pdf that describes possible weld stud failures which might give you some insight.

desertfox

 

[RCMRon]

Desert fox, thanks for the article, and again, thank you everyone for hte great feeback. a lot of my investigation was dependent on that article and other published material including national standards on stud welding. Our failure seem to be classical with respect to the examples shown in this article and others.

I metionned the bolted connection, I have now attached a picture showing this. Keep in mind that the fan blade itself, on which the liners go on and through which the studs go, has 1 1/8" holes and the bolt is only 5/8". The washer, which may be of inferior grade or dimensions, gets cupped which I am sure does not provide any benifits to achieving an equally torque surface. I'll try to get an assembly drawing but our current supplier is not big on those...and a couple of other things.

I went back to check the operating temperature of this fan and it is 200 Deg F (earlier I said Celcius)

We have stopped using impact wrenches and I believe that this is not the origian menthod of attaching the liners. it was modified by a supplier with the use of studs

 

[desertfox]

Hi RCMRon

Your welcome.

I agree that the 1 1/8" hole is not a good idea for use with a 5/8" stud, some of the stud preload is lost when the washer is deformed increasing the possibilty of relative movement between joint materials and stud nut, not to mention the unequal distribution of stud load around the liner stud hole.
Torque is not the best way to generate preload is the error is in the order of +/- 25% possibly more.
The way forward to me, is have the studs analysed at a lab see if they can tell exactly what the problem was, then armed with that information talk to the people welding the studs on the liners and finally get some kind of procedure for tightening the nuts on the studs.

desertfox

 

[unclesyd]

You mentioned in one post that the existing washer was cupping and if so if will be impossible to properly tightened the studs. As mentioned above a thicker washer is called for and if possible a washer twice as thick and hardened. This will give the nut the proper landing area and the effective length of the stud will be a little longer which can't hurt.

I would look at something like the Spiralock nuts for a little more assistance.

http://www.spiralock.com/

 

[byrdj]

it looks like the studs are threaded to the head. I would want a solid shank through the mating plane. given how thin the pieces are, this would require close observation to ensure the studs were stretched before nut bottomed out.

also i would think a snug (press) fit would prevent possible slip between the pieces.