First Guess
Dear mdupont,
in the first place, my request if you could perhaps add a sketch or an excerpt from a GA?!
However,
a) if the swing base mounting is by a rigid coupling re parallel key, you would need an element restricting axial movement?! The torque support / reaction arm as per se can i m e transmit tension / compressive forces only (reaction force of torque & weight), but is in general equipped with some elastic part or is at least flexible in the transversal or angular directions --> and can not support transversal load.
b) Then, there is the point of the characteristics of the broken shaft. How does the crack surface look like? Is the keyway the initial point from where a crack propagates? Did the crack grow, or is the shaft broken in one uniform type of crack surface? --> Sketch / pics would be most helpful!
c) If the reaction arm is not center with the arrangement, pls. cf. to attached sketch. Ultimately, the pulley shaft has to additionally support the reaction force (moment of force) as Mass of drive unit * grav. acc. g * lever l. This must be considered in the shaft calculation, with or without key. This sketch is however only a principle sketch without looking onto the front side / top projections.
d) Last thing, could you perhaps describe the failure mode, in relation to your calculations? Did the shaft break during steady state running or at a instationary operation (braking, acceleration, emergency stop)? The mentioned safety factor, does it relate to statical load, or is this a fatigue strength? How long did it take after first failure to arrive at the next beaking?
Regards
R. ■
All At Sea
As Roland says a GA is necessary. Before and after GA's are essential: although considering the intermediate shaft machining stage the drawing review exercise might be very tedious.
You have managed to involve good old fashioned mixed units (HP) as is the custom with many posts hereabouts. Please try to conform to the internationally accepted system.
there is no indication where the offensive drive is situated; nor why the original set up was replaced, and when.
"We have double drive pulley swing base set up on a stacker/reclaimer ." + "Coupling and shaft have key way." This just reinforces requests for proper GA's, if they are available. ■
John Gateley johngateley@hotmail.com www.the-credible-bulk.com
Drive Shaft Swing Base
Thanks for reply,
Please here the drawing of the GA and photos
A) Its a swing base mounting with rigid coupling. The reaction arm is flexible. The link is a flat bar with 2 pins. It can't hold transversal load.
B) The failure is du to a fatigue strength at key way. Key way is the initial point of crack propagation. Shaft materiel is 4140 HR. First time, we specify to grind key stock edge with a 1/4'' chamfer. It wasn't done. See picture of failure.
C) Reaction arm is not center with arrangement. Swing base arrangement weight 12 000lbs , design calculation 20 000 lbs.
D) It happen during steady state running but some in stationary operation may also increase stress at shaft. Safety factor related to static and fatigue load. Second failure occurred 4 months after first one. First failure was on right side and second on the opposite. After first failure, shaft and pulley were replaced and dia increase at coupling.
We have the same arrangement on both pulley's side.
If more info is required please let me know
Thanks
Regards ■
Second Guess-Worth A Try???
Right. Thanks for the downloads.
There are 2x5.5tonne (about) masses hanging from the shaft with 312mm local offsets.
In the original shaft design with shaft mounted gearboxes the torque arm was probably nearly central and so there would have been less local bending load from the drives' deadweight. Also the shaft mounted unit would have been lighter, but a real pain to remove.
Is it possible to twist and rotate the torque arm attachment to point downwards and then fit another attachment 624mm away on the other side? This would eliminate the extra shaft bending caused by the offset and if you could fabricate a torque arm to restrain the lateral movement and twist on the vertical axis that would also help.
Whatever structural modification is needed it is surely more cost effective than regularly repairing conveyor drives in the conditions you have.
(I am stuck in tropical retirement and I often forget what cold weather is all about. You guys deserve medals.) ■
John Gateley johngateley@hotmail.com www.the-credible-bulk.com
Going Further Into Details
Dear mdupont,
thank you for your input.
So, pls. consider the following as some thoughts from distance, and bear in mind that you might need to involve someone specialized for such job on site = at the material facts.
a) broken shaft crack surface
To my eye, in general it is probably sufficiently sized for the statical load, but due to the dynamical (rotating under bending moment) stresses there's some point to be made. First: Statical and dynamical safety factors --> for the critical section -- will in all probability not give the same value, pls. check. Secondly, the material of the shaft appears to me as a very hard steel, but then probably rather much more hard than tough. Pls check the suitability, esp. considering a sufficient toughness. Thirdly: If this section of the shaft proved now to be susceptible to dynamical bending, pls. check the flat end design of the parallel key. This is, in overlay to bending & torque, a very (very!) influential notch! Dynamic design calc. should show a very high stress concentration in this place. Consider other designs, perhaps extension into less loaded zones, other torque transmitting principles, ... Flat end, its american style, is it?
b) overhung load
As stated previously & above by John, or avoid such load (best practice) or design the shaft adequately, but this would be imo much more laborious. But that is probably not everything, so check the shaft anyway. The crack formation might have been due to some singular overload event, whereas the crack propagation ... and so on and so forth. Have the material / crack surface to be checked by some specialist, if you could reach out to such support.
c) drive design
The drive base appears to be slim, prone to deformation if considering the faraway torque arm support. Did you ever encounter other problems, as fluid cplg leakages, h/s coupling wear or so? However, perhaps you could have that (deformation) checked too, if you're about it?
Looking at the pic of the damaged drive, apart from the corroded state (why that?, pls. forgive my blunt asking..), i'd propose to add a so called shaft breakaway support underneath the gearbox in line with the output shaft. This would be some structure or concrete support just a few inches below lower edge of base, and stopping the drop of the drive effectively + hindering it to fall to the side, causing more damage (ripping of cables..).
It would be interesting to hear from you, what you will do & to which result!
Best of luck, & Regards
R. ■
Some Shared Value?
Hello mdupont,
would you be so kind as to post some info as to where you got with your conv. drive adaptation and how you addressed issues raised?
Thank you in advance
Regards
R. ■
Swingbase Load Analysis
First Observation:
CDI have designed many (>100) swingbase drives up to 3000 kW/drive. All of our designs use a balanced moment approach. You take the reaction of the drive, typically 80% of design capacity ( where most fatigue cycles will be generated) and balance this against the center of gravity to obtain a zero moment at the low speed coupling, both plan view planes. I believe the gearless drive will replace thiese largest size motor connections (>2000 kW) due to the stress concentration problem, for future designs.
Your design does fit this necessity for a balanced moment criteria. First, the reaction point would not be located at the end of the swingbase. Second, the reaction point in plan shows you are off the centerline of the reducer and motor, thereby adding additional bending stress to the coupling. The object is to use the mass of the drive system (motor, reducer, swingbase, et al.) to counter the motor torque reaction to obtain zero bending stress as stated above for both planes in plan view.
I would suspect you will find an unacceptable stress level at the coupling that is leading to your failure, especially with any machine notch stress concentrations. Since you have not provided the coupling machine details, this note is all that can be said. As others have stated, you must also consider the momentary overload due to acceleration or braking. CDI does not use keyways above 200 mm diameter shafts. There are a number of options to proper rigid coupling designs which do not use keyways. Keyways are a cheap method of torque transmission that requires a larger shaft to compensate for the keyway stress riser, separate from any shaft diameter step configuration.
Another point I did not observe is the connection between swingbase and earth. Some have tried to not use a double ended spherical connection, allowing the torrque are to freely move to relieve the precession displacement developed by inaccuracies in manufacturing of the coupling halves. This can also lead to high local bending forces that can lead to shaft failure at the coupling.
To also reinforce other comments, the fatigue life must also consider your materials of construction. ■
Observing The Sketch
Seeing the Sketch, I observe your location of the torque are exacerbates the bending stress with the "L" value. You are going in the wrong direction. You must generate a moment diagram with the drive reaction force pushing up at the torque arm and the consider the gravity force trying to balance this condition, taking into account the pillowblock bearing position.
Think about the upward force of the drive torque resolved into a force at your torque arm. Then locate this torque arm such that the downward center of gravity causes zero bending moment at the center of the rigid coupling, according to the set of drive steady-state and momentary conditions. You can even do better by counting the loading cycles for each class of loads and do a fatigue analysis for all these expected classes of loads using "Miner Rule" or other often used procedures. ■
For Free But (Hopefully) Not In Vain
There's so much good value here from John and Mr. Nordell, it would be so good to know whether it was of any help on this here issue. I dunno, could be it's over there a so very tough environment rel. to contract / staff / human factor that one cannot give away nothing, but it's nevertheless the feedback that makes us = all advance. But the feedback. I don't like the feel the like of a preacher to the benighted heathen, it's human to talk back and be it to say that the shot went totally astray. Or to be given a chance to readjust. Or to give others the chance to avoid similar.
Dead ends are a horror, from an engineering point of view.
A good weekend to all
R. ■
Drive Shaft Swing Base
Good day experts,
We have double drive pulley swing base set up on a stacker/reclaimer . Previously, the gearbox was shaft mount. The system was replaced last year by a gearbox with rigid coupling. Since, the shaft broke twice at pulley's coupling. All calculations show a safety factor of 3 on shaft a full torque (750hp).
First time, we remove de step at coupling to avoid force concentration (200 mm straight: originally step down shaft to coupling 200mm/190mm). Coupling and shaft have key way. The shaft broke a second time at the same location.
Is keyless coupling will solve the problem and/or we have to look for something else?
Also, we are look at swing base design. The reaction arm of swing base is not center with the arrangement (312mm) in Z axis (because of site physical restriction). Is this can be a cause? Knowing the distance between pulley axis and reaction pin is 3600mm.
Thanks to all for your help ■