Crack detection 3

[WardHollowayPE]

I am looking for a coating (paint like) that can be applied to any part of a machine. When the part cracks the coating will make it blaringly obvious. The part can then be replaced before total failure.

 

[GregLocock]

Plaster of paris? Cheers

Greg Locock

 

[butelja]

I think Magnaflux Corporation used to have what you are after. This is state of the art 1940's technology. This comes from their web site:

"1942
Aeronautical engineer Greer Ellis is granted a patent for Stesscoat®, a brittle coating used in quantitative stress analysis. During that same year, Continental Motors uses Stresscoat® for design/weight reduction and gets published in LIFE magazine giving widespread exposure to MAGNAFLUX."

Unfortunately, I don't think they make this coating any more. You can try

http://www.magnaflux-online.com/

to get a contact. Maybe they have some Stresscoat in a warehouse gathering dust.

 

[Biggadike]

The stress sensitive varnish is what you are describing. The problem it that it needs to be used in a scientific manner (progressive loading) to determine the point of maximum stress. That isn't the same as a 'paint it and forget it' solution which you seem to be looking for.

You seem to be suggesting that your parts are cracking in normal service. This suggests they are overstressed of manufctured with crack-like defects which grow during service. Only something like an ultrasound scan will show these up. I know of no way of getting a quick visual indication of when a part has cracked beyond its limits.

 

[Ron]

Materials such as this are still available and the method is still applicable. Check with MicroMeasurements in Raleigh, NC. I believe they supply these.

 

[WardHollowayPE]

Thanks for all the reply's
Just to elaborate on the problem, we produce a high speed production machine mostly made out of steel. In order to achieve higher production rates, I am going to re-design some parts out of aluminum. Because these parts experience high cyclic loading, the aluminum will eventually fail. The plan is a two part construction, one smaller than the other. The smaller part should fail first giving indication of the failure.
What do you think?

 

[MadMango]

Sounds like you are designing for the failure, instead of improving the component so it won't fail.

If you have the ability, you might be better off installing a vibration sensor connected to an alarm (audible or visual). Under normal operation you would know the standard frequency, and if the component started to fail, the frequency would change.

If that sounds too complicated, you might want to check out this coating that NASA uses, it's $20 sqft.

http://www.isa.org/journals/intech/brief/1,1161,384,00.html

"Happy the Hare at morning for she is ignorant to the Hunter's waking thoughts."

 

[TheBlacksmith]

I have been reading this post and its replies for days. I am still trying to come to grips with redesigning the parts and knowing they are going to fail. My usual rant- run the numbers. Are the profits of increased production going to offset the downtime when the parts fail? Does said failure have the possibility of damaging other parts, increasing the costs and downtime? What about using another material that has both the lighter weight and high fatigue life, titanium, some of the new composites, etc. Maybe the lack of failures will offset the costs. Maybe you're being driven by the bean-counters, but planning on the failure to the extent that you're looking for a coating that makes it obvious...

Blacksmith

 

[WardHollowayPE]

I am planning for failure, sounds strange everytime I say it. The speed of the machine is limited to the amount of weight it has to through around. So I plan to eliminate that weight. I can calculate the life of the aluminum under cyclic loading, but I can't tell the customer to stop the machine ten years later. I get calls from customers with machines over 30 years old. They just keep on rebuilding them, or fix it when it breaks. A planned and obvious failure will force the customer to call up and order a replacement part.

 

[cranekiran]

Well!,
To me it sounds like (I have been following this thread)the choice very similar to what we have between a replaceable rubber seated butterfly valve where the rubber sleeve is replaceable and the integral rubber seated butterfly valve where the rubber seat is totally moulded with the body.
I guess whatever the methods are,It is best left to the customer to decide what kind of product suits him best.In the case of your product if it is the aluminium part which you want it to be replaceable then you have to first convince the customer of the economical benifits of the new design and also you have to be carefull not to have more than the required numbers of the replaceble parts because in the end the customer should feel confident about your new designed machine.

 

[3DDon]

Ward,

I have some ejection seat background. All ejection seat systems have at least one high pressure braided hose. One the ACES II (as found in the F-16, F-15, A-10, B-1, and B-2 aircraft), on either one or both sides of the seat, these high pressure hoses can be found. The fitting on the end are screwed all threaded together. After these fittings are connected and torqued, we use to apply a thin stripe of bright yellow enamel paint across the fittings. If they were to loosen, the pint would crack and flake off. It was real easy to see if something moved.

Is it possible that this type of method could be used?
Don Shoebridge
Sr. Product Developement Engineer

http://www.geocities.com/donshoebridge

 

[MadMango]

WardHolloway wrote- "I can calculate the life of the aluminum under cyclic loading, but I can't tell the customer to stop the machine ten years later."

Why can't you tell the customer? If this is the type of high-speed assembly machine that I am thinking, wouldn't it come with a service/owner manual? Wouldn't this manual suggest a standard maintenance schedule so the customer could keep this expensive machine working at an optimum level?

If so, you could tell the customer to check this "failure part" on a bi-annual or annual basis, looking for signs of fatigue. Sounds like you are designing it to fail anyway, so it's already a repair/replacement part to begin with. "Happy the Hare at morning for she is ignorant to the Hunter's waking thoughts."

 

[WardHollowayPE]

Can I really rely on a customer to check a failure prone part before catastrophic failure? A failure at this rate of operation could cause serious damage. I think its better to have an obvious mechanism in place to indicate a near failure before the damage is done. Even better would be to make the machine inoperatble before complete breakdown. A two step failure so to speak.

 

[cranekiran]

I really do not understand your point,wardholloway.As you have pointed I think your machine must be a continuosly operated high production tool(correct me if I am wrong),Then your machine has to have a maintanence rest period just like others in shop floor.Every industry at present(unlike the dark ages) has a maintanence dept. which follows a test log book for each machine for a set time period.They (maintanence dept)have to update the machine performance based on the current standards(ISO)on their files strictly on a daily basis or on whatever the machine rest time period is;so pray tell me how will the operator fail to check the machine?
I still strongly feel that at some point you have to trust your customer.

 

[ProEpro]

Were the previous parts designed for an infinate fatigue life? How long did they last? Are there any other parts that wear and need replacing like bearings? If you integrated them or at least sold them together there would be a better chance the parts would get replaced.

It sounds like you are looking for something to act as a mechanical fuse. It is allways a good idea to think about how something would fail and the results. This is often called a failure mode analysis or FMEA and there are people that make a serious science out of it.

If you use the wrong paint it will hide cracks instead of showing them.

Good luck and please tell us what your final decision is and the outcome.

 

[WardHollowayPE]

I like the term "mechanical fuse" thats just what Im looking for. The original part is made from cast iron, very heavy. In operates at 400 cycles per minute, the operation can vary from 8 to 24 hours a day. The operation resembles a crank and slide. The part I am replacing is the connecting rod, next I will work on the slide. Because it is made of cast, it has a infinite cycle life, and is limited to the 400 cycles per minute due to weight. I plan to at least double this output. My design' as of now' looks like a two piece construction of aluminum. One part is smaller and may be notch for failure. This will be the fuse. No matter what the life span, I worry about a failure at high speed. A one piece construction may fail and cause serious damage. I like the two piece mechanical fuse idea.
Thanks to all for your input.

 

[GregLocock]

Hmm. If long life is your objective then I think you are making a mistake in going to aluminium, since, eventually, it will fatigue. For a conn rod you are basically looking for some sort of stress/density optimisation, so I'd have thought a decent cast steel or even a forged part (say axle steel) would equal the weight of an aluminium part, and could be designed for infinite life.

Sorry if I'm running over old ground for you, but I'm a founder member of the 'steel is often as light as aluminium, when all things are considered' club.

Or you could invoke the grand tradition of copying what everybody else does: what are automotive conrods made of? Cheers

Greg Locock

 

[cranekiran]

Allright! If you have already decided to go ahead with your replacement idea,How about using a Bearing or Self lubricating bushes?They can be easily replaced. See,the replacement itself should not be too complicated.Anyway,whatever you decide My best wishes to you for your design.

 

[Hush]

How about designing the part so that a non-critical component fails before the life cycle of your main component is approached, requiring the entire replacement of the part. I've seen this approach used for lifting arms on farm implements where a journal bearing will fail prior to the arm failing and killing someone. Ticks the farmers off especially when it happens mid harvest but probably has saved a few lives.

Another device I've seen is a shear wire, a thin wire that will break at the first fracture or over-stress signalling an alarm.

 

Good grief! You want low inertia and high cycle life, go to composites!!! Done properly, they can be reasonably priced, light and - you migh not realize this - have almost infinite fatigue life! The fatigue life is something I find most people have a counter view of. They think it will fail sooner than steel, as opposed to the opposite when well designed. Corvette springs are fiberglas or other composite now. A big hunky part like a con rod on a press (getting close?) would be a sinch to make well and BEEFY and light. Matrix "plastic" choice is as important as fiber choice. Compression molding would be a natural. Bearing journals could be finished in molding, helping with the finished part cost. Bearing races could be molded IN, making the whole mess quite simple.

If you think aluminum will fail, you need to get familiar with your crack length at failure - you'll find that a lot of stiff materials are much more ugly about this than mild steel. What loads are you producing that get within a few percent of your member's strength? Your likely failure modes are around the journals, which fiber directionality and concentration would help greatly!!!

Other than that, look again at steel, or properly design an aluminum part - not just knock off a steel part.