broken studs

huangyhg

broken studs
i have a problem with st3174-021 studs  continuously breaking while being installed.  the part they are being installed on is an aluminium casting made of ams 4219 that has been chromic acid anodized as per ams 2470 on exposed outer surfaces, and chemically treated as per ams 2474 "d" ,with immersion in actan, on internal mating surfaces. the part has twentyfour 1/4-20 threaded holes distributed over both types of treated surface that recieve st3174-021 studs. the holes were milled using a tian coated gh3 tap. the problem comes after anodize & chem treat.  assy of the stud requires a wind down torque of 25 - 45 inlbs, but the actual torque required is up around 100 inlbs and more, at which point the stud breaks and additional machining time is needed to remove it/them. an expensive proposition at best. i've tried re-taping the  holes before stud installation and that worked, but it's not even close to an option. re-tapping by hand is too time con$uming, and what's more, it removes the anodize/chem trt. does anyone have any ideas on how to rectify this problem?  
eng-tips forums is member supported.
oil the studs?
mike halloran
pembroke pines, fl, usa
i did try that, but there was no marked improvement. in fact, even once the hole is re-tapped, oil *must* be used to install the stud.  once the anodize is removed, the friction between the sst stud and th al  causes galling, the stud breaks, and i'm back to square one.
then i'd use never-seez.  i'd probably use it anyway.
but no improvement with oil suggests that the holes are just flat too tight, and are not being tapped slightly oversize before finishing to compensate for the thickness of the anodize.
the other choice is to tap the holes to the correct size, then mask them so they won't be anodized.
mike halloran
pembroke pines, fl, usa
i'd love to be able to use never-seez, but it's not an option. this item is part of the fuel delivery system for a turbine engine. to prevent any contamination, the only oil the client permits for lubrication is the fuel that will be passing through it. also, according to the drawing, treatment of the holes is *required*.  they only accepted the part with the re-tapped holes for testing purposes.  
i had wondered about the chromic anodize decreasing the id of the holes, but a colleage insisted  that it does not add any amount of significance. he insisted that only a hard-anodize process would add enough material to cause problems, and that is not the case. does anyone have any knowledge to the contrary? if so, how much material am i looking at?  i'll need to know what size of tap to order oversize in order to compensate.  this may be the avenue i was looking for if in fact the chromic anodize adds material.
it's been my understanding that anodize typically adds a little, certainly enough to screw up a tight threaded joint, but i think the amount is dependent on the chemistry, the time, and the current density.  only your anodizer would know for sure.
warming the part is a possibility.  haven't tried that yet, i will give it a go today & see what happens, but chilling the studs seems pointless.  we routinly use the chill method, either using dry ice or liquid nitrogen, for assembling other components that are usually friction fit. in my experience, you only get about 4 to 7 seconds of assembly time on parts that small.  they just heat up too quickly.  the portion of the stud that goes into the threaded hole is typically 1/2 an inch in length. to wind them in by hand using a torque wrench takes too long for the chilling to be of any benifit.  if we had an airtool or something similar, it might work, but it would also have to have some sort of clutch system that would slip once the desired torque was reached. the stick-out portion of the stud that is used to mate this part to the other parts of the fuel system is also important. the height of the stud from the face of the threaded hole to the top of the stud has a 0.020" tolerance, so the tool also needs to have depth control.  the tool would also have to be chilled before and during use to buy a little more time for assembly.  using a room temperature tool would only heat the stud faster, leaving no benifit from chilling.
still possible, but i'd need to find a tool that will work at very low temp.  i will try warming the part today though and let you know what happens.  thanks.
even a small thickness increase due to anodizing can have a huge effect on threaded joints.  the thread geometry magnifies the effects of coating thickness.
for example, say a bar gets immersed in a coating.  the coating thickness is 1 unit, but the diameter change is 2 units (2 surfaces).  for a thread, a coating thickness of 1 unit changes the pitch diameter by 4 units.  this can be a major contributor to fit problems.
i think you should get your parts checked dimensionally.
regards,
cory
hi
have a look at this site:-
first, i want to thank everyone for answering this post, in spite of the fact that it was posted in the wrong category. i should have posted it in the paint/coatings forum, but i didn't know any better at the time.  after much searching, reading and helpful comments from this site we are now on the road to a solution. a test piece of the same aluminium type has been ordered from the casting foundry, as well as a selection of oversized taps from gh4 to gh9. the plan is to drill & tap 10 test holes from each o/s tap, send the test piece to be anodized and then install studs into each hole to determine which tap size gives the best results.  a realatively simple solution, once the source of the problem is known.  if anyone is curious as to the results, let me know & i will get back to you.
thanx again to everyone.
please post the results right here for all our lurkers to read.
thanks.
mike halloran
pembroke pines, fl, usa
i don't have any experience w/ chromic anodizing, but in standard sulfuric (type-ii) anodizing, we'd generally have a coating thickness of .7-1 mil. however, our net gain in thickness was about negative .3-.5 mil, or about half the coating thickness. hence, we still lost material as we grew the oxide coating from the parent material.
while we were only making decorative trinkets, i'd guess the dynamics of the type i and type ii processes are similar.
update:  the solution will not be here immediately.  we have a 6 to 8 week lead time on the test material. i will post as soon as we have results.
thanx
just noticed this thread;
did you double check your thread call-outs?
using cut threads for the female, and rolled threads for the studs?
could have an interference with that.
thanks for the comment monkeydog. it's something i never even considered.  i will look into that. i'm not sure how the studs are manufactured. the studs are st3174-021, and are supplied by the client.  it's a strange system that i don't understand. the client sends us the casting and all the parts used to assemble it after machining.  the weird part is that they bill us for all the materials & then we bill it all back to them when it's complete. (back to topic). we do have specs on the dimensions of the studs. i know that the major & minor diameters both have a 0.005" tolerance, which is also a problem.  the base of the stud that is being installed is 1/4 - 20, the exposed part of the stud used to mate the parts is 10-32.  the 10-32 threads are very clean visually, and look as though they are cut. the base however is very sloppy looking on some studs, and somewhat clean on others, but not nearly as visually clean as the 10-32 section.  in most instances that i know of, threads such as these are mass produced on screw machines, but i have no clue which process they use. i'm guessing that there are multiple suppliers as well. it could very well be that any one supplier could use either method. is there an easy way to tell which method was used without tracing the parts back to the manufacterer & asking? if the studs are rolled, would switching to a roll tap instead of a ground tap be any help?
rolled threads:
- typically have a smooth runout at the transition to the shank.
- may have incomplete or slightly grooved crests, where material was plastically displaced upward from the blank.
- have a shank smaller than the thread o.d., because the blank is smaller than the finished diameter over the crests.
- may have a cupped surface at the distal end, again due to plastic deformation pushing material off the end.
cut threads:
- stop abruptly if single- pointed, or in 3..4 distinct steps if cut with a die, always with a fracture surface where the chip(s) broke off.
- have a shank diameter equal to finished o.d., or slightly larger.
they're very distinct from each other under modest magnification.
screws are almost never made on screw machines, since the invention of cold headers ... but because of the step in o.d., your studs would be difficult parts either way.
mike halloran
pembroke pines, fl, usa
thanks mike.  i have,in fact, looked at these studs with a 10x loup. if i had to bet, i would say that, by your description, the 1/4 - 20 section was rolled, and the 10-32 section was cut. but i'm guessing.  i have been googling "st fasteners" in different fashions to try to find the specs on the manufacturing process, but so far have come up empty. any site you know of would be helpful.
i used to have an old tolerance study report, that discussed the potential interference when you combine controlled root radius theads (mil-s-8879) with cut threads (mil-s-7742), i cannot seem to find that report, may be a hard copy back at work.
anyone can replicate the results, just take your call-out of threads and draw them out (10x) worst case tolerance, and overlay the drawings.
this is probably not an option for you, but on some components i have worked on, i have found helical coil threaded inserts as part of the original design / assembly. the inserts could be easily installed in your treated part then the studs threaded into the "helicoil".
thank you for the post fieldteam. i agree, helicoils would be a great choice.  we commonly use them in our shop for different jobs, but unfortunately, the design team on the other end does not seem to think it's necessary. helicoils add cost to the finished part which they do not want to even consider. last time i checked, the cost of an st3031-43 helicoil was about 14 cents. i know, that the st3031-43 is for a 10-32 hole and not a 1/4-20, but it's the first one that came to mind and i didn't figure that the cost diff would be all that much, but it was still not an option. unfortunately, i'm not sure if i will ever be able to post the results now. as it turns out, i was contacted by another company with an offer of employment that i just couldn't turn down. they matched all my current benifits for the company that i work with now, and then added more money, more interesting challenges, and the kicker (for me), paid learning! they are going to pay for any studies that i wish to take, as long as it pertains to my field of work. i already have a list of courses i want to take and they are already preparing the paperwork.  i still have contacts at the old place, and i am in the midst of completing my obligatory 2 weeks notice, but i can not wait to start the new job. i have made a few good friends after 7 years at the old place, so it's possible that i might be able to pry some info out of them on how it is/was eventually solved, but i can't make any promises. i will do what i can.  i am still curious myself.  i want to thank everyone who posted a reply, and hopefully, i will still be able to post the final solution. we shall see what happens.