Estimating fatigue on monorail track 1

[bsmet95]

My company manufactures cranes and monorails using patented track. One of our installations has thousands of feet of track and hundreds of 2-wheel trolleys. See the attachment. The customer wants numbers on fatigue of the track. We know the wheel loading and number of trolleys per day which operate on the rail. Is there any way to do a fatigue analysis on the track?

Thanks.

 

[IRstuff]

Wouldn't you start by looking at the maximum deflections exerted by the loads, and determine the total number of cycles over the design life of the track?

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[UcfSE]

Yes, you or someone can assess the track for fatigue. Your question is a little broad to provide specific answers. You need to understand how much you are willing to spend or how much the customer is willing to pay for the information.

 

[rb1957]

rather than maximum deflections I think you need "typical" deflections from "typical" loads (sure maybe a fraction of the maximum).

The point is to start with a stress analysis of the track as this customer has implemented it (supports etc). Then how this customer typically uses it (typical load 80% maximum, typical usage 200/day). You can side-step the daily usage (frequency of the loads) if your track stress for the fatigue (typical) load is less than the endurance limit, or so small as to be negligible.

This stress analysis can be a daunting exercise (as you have to go down to the smallest details).

If doing this is not your normal business practice then either
1) charge him for the work, or
2) politely decline.

another day in paradise, or is paradise one day closer ?

 

[LiftDivergence]

Honestly, this sounds to me like something you're going to need to research heavily. I wouldn't underestimate the complexity of this problem.

Based on your diagram I think this is more of an issue in the realm of contact mechanics where you will have Hertzian contact stresses.

Another question - are there any junctions or discontinuities in the track? What does the global view look like? Be aware of any stress risers that could shift the focus of your fatigue analysis.

Maybe start with these:
K.L. Johnson - Contact Mechanics
S.L. Grassie - Mechanics and Fatigue in Wheel/Rail Contact 1st Edition

I'm sure you can dig up more references on exactly this problem.

Keep em' Flying
//Fight Corrosion!

 

[dik]

...and my experience is that it is 'really tough to do' and your estimate may have no connection to reality.

Dik

 

[bsmet95]

Thanks , everyone, for your input. I'll inform my boss and see how he wants to proceed.

 

[rb1957]

on the face of it flange bending of the track could be the critical location, as the trolley passes any point. There'd be an effect (possibly secondary) from the track bending as a beam between it's supports ... which would complicate things significantly. There could easily be some small detail hidden in the structure (in the trolley, in the track supports) that has a short fatigue life (under whatever cyclic loads it sees).

another day in paradise, or is paradise one day closer ?

 

[dhengr]

Bsmet95:
Good clean detailing, welding, machining, fab. and assembly, and real engineering attention to a thorough structural analysis of the entire structural system with due attention to stress levels, and combined stresses has much more to do with providing a fatigue resistant product than 500 pages of mathematical crap and verbiage, which bears little resemblance to all the possible real-world conditions. You should have considerable knowledge and experience about which design aspects/approaches and details work well over the long term. You should have a good handle on the load/cycle conditions for the installation. There will be parts of the system which see only a few cycles per day, and there will be some locations, with multiple handling actions per hour, at a particularly busy plant station. So, tabulate a list of various cycle v. loading conditions, to show that you have considered that aspect. There will be details and locations on your particular system which are higher stress, more fatigue prone, and you should pay some attention to them, and show it in your calcs. and presentation. Examples might be…, mid span moment plus bot. flg. wheel load cantilever bending, negative moment at supports, beam/rail stability conditions, bot. flg. splice details and wheel load interaction, to name a few.

This report or explanation of your company’s thinking/philosophy on fatigue should be a bit of a tutorial on fatigue considerations in the real world design/build environment. It should also touch on your good history of no fatigue problems with your product over the years, under some severe usage conditions. It should touch on your QC program to minimize fatigue prone details and conditions, etc. It should comment on the detrimental affect of over loading and serious mis-use by the customer’s plant people and what factor of safety you apply to help cover this.

There are a number of lab approaches to studying fatigue problems and life on simple elements under controlled lab conditions. And, these shed some light on particular materials and simple detail conditions. There are darn few good ones for real world fab. and details, except to fab. 1000 of those details and test them to failure and start to draw some conclusions about the good and bad aspects of each detail. That has been done on many typical details over the years, so we have some history and understanding of what constitutes a good detail and what kinds of stress raisers cause serious problems. You should study these as they apply to your product, and get to know a good detail when you see it.

 

[dik]

As part of the problem you are looking at, I've attached a part of my project notes for crane related projects:

CRANE SHALL BE DESIGNED FOR A DUTY CYCLE OF 10 LIFTS PER DAY, FOR 250 DAYS / YEAR FOR A PERIOD OF 20 YEARS (50,000 LIFTING OPERATIONS)

% CAPACITY % USEAGE No. OF LIFTS
100 5 2500
80 50 25000
50 30 15000
20 10 5000
10 5 2500

This has an impact on the fatigue life of the structure, and one of the reasons it's so difficult to determine a 'life expectancy'... inspection cycles are essential and more frequent as the system ages. Stuff can still be missed.

Formatting doesn't work so well... matter of spacing the text out a bit.

Dik

 

[GregLocock]

I'd have thought railway engineers would have a good handle on this stuff.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

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

 

[rb1957]

maybe, but I see a railway track (on sleepers) being much more uniformly supported, and the load isn't really on the the flange.

another day in paradise, or is paradise one day closer ?

 

[Nigel Linden]

I would take a section of the track from one support to the next, and assume a point load in the middle, on either side. Model it using ANSYS, ABAQUS or equivalent. You can use symmetry constraints to reduce the size of the model, but it should be fairly simple. Then you can use nCode's fatigue analysis if ANSYS, or FE-Safe if you use ABAQUS. This is a fairly simple problem, and so you can use a stress-based approach. The duty cycle should be straight forward. As the wheel traverses the track it will effectively create one cycle of the point load at any given location. Fatigue is all about cycle reversals, and so you don't need to worry about the effect of the wheel rolling along the track, because that will all be a part of the same cycle.

If you already know the maximum stress, you can avoid much of this and use an S-N curve for the material to estimate cycles to failure.

Let me know if you need help.

re-test.com

 

[MikeHalloran]

Right after college, I worked in a car parts factory, that had dozens of monorail loops carrying parts from one station to another on linked 2-wheel trolleys, with quite a few steep grades up and down (e.g. to pass over another loop, machine, or aisle) and quite a few lateral turns of short radius, laid out to suit the material flow in the factory.

I'm sure the conveyors were well engineered and all, but they still broke regularly, especially on new-ish lines like mine, where they had installed yet another new conveyor design because the older designs either had problems or were perceived as too expensive because they didn't break often.

I damn near got killed by one where the inter-trolley link broke at the top of a hill, and hundreds of axle tube assemblies flowed down and were thrown off their hooks. The cascade stopped after a few seconds, when the drive overloaded and shut down the loop, or maybe somebody whacked an e-stop, but either way it was scary as hell. I was standing near the growing pile, stuck like a deer in the headlights. I owe my life to an hourly worker who shoved me out of harm's way just in time.

I just wanted to comment on the appropriate number for 'lifts per day'.
In our case, we were selling about 2 million cars a year.
With 200 working days per year, that works out to 10,000 carsets of parts per day.
Use numbers like that for serious conveyors subject to fatigue.




Mike Halloran
Pembroke Pines, FL, USA