Heat shrink fitting and crankshaft balancing.

[picclock]

Hi
I intend to make a crankshaft by using heat to shrink fit the parts together. Does anyone know of a resource or set of rules whereby I can calculate the correct sizes and temperatures to shrink fit the parts?

To explain more clearly, the main shaft of the crankshaft assembly will be 10mm silver steel. So to heatshrink the crankweb, 8mm thick steel, to the main shaft I will need to make a hole that is smaller than 10mm, then heat the crankweb until the metal expands enough to force the 10mm steel into the hole. So I need to know how much smaller than the 10mm to make the crankweb hole and what temperatures I will need to heat the webs to.

This seems to me to be the best way of making a truly rigid and accurate  assembly. I can leave the  main shaft as 1 long piece  and shrink on the crankwebs and big ends. Then I can cut out the parts of the main shaft that are no longer needed. Because its all assembled with the shaft in one piece the alignment should be spot on.

The other thing I need to know is about how to balance the crankshaft for minimal vibration. The amount of metal left on the crankwebs as a counterbalance should be defined by a formula or somthing similar. If anyone can help with any info on this I could possibly stretch to a virtual beer. 

Many thanks

picclock

[John Rudd]

I would have thought the best way to make a crankshaft would be out of one piece turned up in the lathe...less effort than trying to piece together a load of hot bits and then beating them into submission to get the bits to line up...

Mebbe someone else can chime in with a better way...

[picclock]

Hi John. You are correct but :

1. I want to fit needle roller bearings to the assembly (mains and big ends), which makes it tricky to get them on :-).
2. I reckon I would need a 40mm piece of steel to turn down which is a HUGE (for me) amount of work.    Also I would find it hard to get the required level of finish for said roller bearings.

I could perhaps turn down sections but I think the heat shrink way should give good results with less effort. Also I can harden the silver steel mainshaft and big end pieces before assembly. Cutting a hardened shaft is not difficult with an inox grinding wheel cutter. I considered 'Bogs' puddleduck engine method but concluded it would be too floppy. I am hoping to get some serious power out of this thing so it needs a good stiff crank. I have some heavy duty arc welding kit if it comes to it, but the shrink idea would probably be stronger and need less cleanup. 

Using the 'shrink' method I can turn down the shaft where it needs to be removed  (by 2mm or so) and supply a lead in for the cankweb to start on. Once it's done removing the unwanted bits is easy. I'm only making a two cylinder engine so it should be fairly easy to do (famous last words).

I also considered tapering the shafts and holes, applying a lot of pressure, then drilling a hole and using a steel pin to lock it, or threading the parts with a small shoulder and locking them with a through pin.
 
Many thanks

picclock

[rleete]

Make lands on the shaft, and press fit the parts on.  .001-.002 interference and red locktite will hold up to an amazing amount of force.

[picclock]

Hi rleete

We may have a communication problem. When you say lands do you mean flats like if you looked at the shaft end on, it's shaped like a D ?.  Not a problem to do but how do you cut a hole thats D shaped ?

Let me know if I got it wrong.

Many thanks

picclock

[bp]

No I think he means a shoulder.  On your 10.00mm dia shaft, at each end for a length of the web thickness (8mm?) reduce the diameter to (say) 9.80mm, in the web generate a hole 9.78/9.79mm diameter, preferably by reaming, a good finish anyway.  Think really hard to figure out EXACTLY how you will assemble everything, maintaining squareness is vital and then practise how you will assemble everything when one bit is REALLY hot, bear in mind that considerable force will be required which is why its called an interference fit.  Some sort of vice or press is really good, a hammer would be very much a last resort.  Warm up the web(domestic oven hot as possible for half an hour), cool down the shaft (domestic fridge for half an hour) and press one onto t'other using your considered and practised process.
cheers
Bill Pudney
Adelaide, Australia

[Country Bubba]

If'in a dumb ole country boy can jump in on this thread,  back many years ago, I used to work on "American Indian" motorcycles and the crank pins were on a taper and had threaded ends.  You would assemble the rods with the roller bearings and then use a stand with centers to get the two crankshaft ends to line up properly. I think Harley also used this type of thing.

Country bubba

 

[bp]

yup ole country boy, not only Injuns and the other one, but in the UK some of the biggish singles (350 & 500) had a similar set up.  Getting a fine taper to pull up EXACTLY where you want it too might be a challenge though, and then having enough space to use a (special) nut to stop it all falling apart.  From memory most of the UK manufacturers used a parallel pin, with a very accurately ground diameter, going into an accurately ground bore, all held together with a very fine threaded nut.  BSA on their Gold Stars even used rivets to secure the flywheels to the outer bits of the crankshaft, instead of a nut, Victorian engineering at its best!!
cheers
Bill Pudney
Adelaide, Australia

[picclock]

Hi Bill

That makes much more sense. I was planning to heat the webs to 5-600C with a torch then use an old scrap drill press (once I got my mill its never been used) to push the shaft squarely into the hole.  

OK I've done some calculations that indicate that the difference in size due to expansion, at 600C is only 2.9 thou, thats 0.074mm. I think that may be achieveable. I can always make the shaft ends +2.5 thou then up the temperature to 800 for clearance.  I think its doable for a shaft into a web. But to get the webs accurately aligned with bearings on certainly sounds more challenging. I can fit two shafts into 1 web x2 for the ends with no special alignment. The centre portion of the main shaft will need two webs fitting which must be at 180 degrees and have a needle roller fitted inbetween. Then I need to fit the end assemblies with needle rollers parallel to their respective central webs.  

My main concern is the web alignment, with bearing damage or contamination next. I may have to rethink this.

Other possibilities that come to mind are a tight press fit with an 8mm csk hex screw in the end, or welding (yuk).

Many thanks

picclock

[John Rudd]

When I rebuilt cranks for my motorbike many years ago, I had access to a 20 ton press...
Even with the tight tolerances it took nearly 10ton   to reassemble and everything was absolutely clean...

Then I need to fit the end assemblies with needle rollers parallel to their respective central webs. 

Are you using thrust washers as well? You will need to allow for some side clearance....

Good luck with it..

[bob ward]

I'm late to this thread, but FWIW I've been trying to achieve heat shrink fits on 10mm dia items (there are other ways to do what I'm trying to do but I've not played with heat shrinks before so I thought I'd give it a go) and have abandoned the idea.

Achieving the required dimensions for the male and female was not a problem, neither was heating the items to the correct temperature. Where it all came (un)stuck was in fitting the parts together, in 12mm deep holes I could never get the male part in more then 8mm before it jammed, and jammed forever.

What I think was happening was that given the tiny air gap between the 2 parts and the small mass of the male part, radiant heat from the female part was almost instantaneously growing the male part. I'm not saying you couldn't make heat shrinking work on something this small, but you would need a setup where you plant the male into the female like an absolute rocket.

[Lew_Merrick_PE]

It appears the original post has been answered with do it a different way.  However, the questions about the strength of an interference fit and how to calculate expansion of part when heated was not answered.

Both Machinery's Handbook (and the imitators) and Shigley's Machine Design have sections on calculating the strength of an interference fit joint.  You are basically creating a localized yielding of the two parts which creates a preload on the joint.  There is a section in Shigley's Mechanical Engineering Design that has exercises in making these calculations.

You can find exercises on thermal expansion in most mechanical design engineering texts.  Shigley's Mechanical Engineering Design is one of the best.  You end up needing to know the coefficient of thermal expansion (CTE) for your materials.  http://www.matweb.com has the most complete and verified listing of such properties.  The CTE has units of length/length/degree of temperature (i.e. in/in/°F or m/m/°C or some variation thereof).  Be sure your units are correct.

I hope this helps.

[madman]

Crank Pinning... or tig weld after Alignment Completed. If you drill small hole bisecting the crank and crank pin and slight press a small pin it will Not move, Then bonk the hole with a centre punch (120 degrees apart) to keep pin in place. Ive done this on Cranks  for my Motorbike engines.  As for balancing I do know People make up bobweight assemblies (adjustable for different weight of Lead shot cylinder hollow with threaded end cap and fill with different weight lead shot until ideal operating rpm range and smooth running is achieved)  and factor in a percentage of recipricating weight (like a 68 percent factor) then get a guy who knows what hes doing to balance it. I would also like to know how to balance one. I do know that years ago on Honda 750 engines we modded, we hacksawed the crank throws  and filed sanded smooth. It never seemed to effect the RPM at high RPMs. We built 1000 cc plus alcohol engines from the Base 750 Hohda SOHC piles we had. I was always suroprised that our  Crankshaft Hackery didnt shalke it apart at HIGH RPM where we always ran them.

[picclock]

Hi Madmen (welcome to the forum)

That sounds like an easy way to do the job. I'm finding it a bit difficult to visualise the method you wrote about. Sketch below is my interpretation, although inserting the pin from the top into a blind hole may be better (but only if you do not want to remove it  ).

Peening over the edge of the hole sounds like a foolproof way to prevent pin escape, although if I go this route I may try to find some small interior circlips to retain the pin.

It would be nice to be able to calculate the balance mass required as its a lot easier to change the size of the crankshaft webs during manufacture, so getting a close approximation would be useful. The lead shot idea seems good to me for fine tuning.

many thanks

picclock






 

[RichardShute]

There are plenty of (mostly very technical) texts on machine balancing. There are a number of considerations, primary balance and secondary being the fundamentals.
Taking a simplistic view of what I can recall of a subject I've not studied for 30 years, if you have 100% primary balance, the up going mass of the crank, conrod, piston assy is matched by an equal down going balance weight mass.

However, the bad news is that at any point other than straight up or down, the balance mass is flying out sideways whereas the pison and most of teh con-rod are moving vertically resulting in major grief, secondary imbalance. With multi cylinder engines you can mitigate the effect a lot by balancing one cylinder's secondary against another's 180 deg apart. This is still not by any means perfect since the two secondaries are not acting at the same point so you have a couple which also rotates, generally flailing the whole thing about.

Vee configuraion engines can capitalise on having short cranks whcih helps and also using secondary balance from more than one cylinder to balance primary balance of another.

If you want to get into it in more detail, I'll try to dig up some reading.... hope you enjoy trig.

Richard

[madman]

Picclock no thats NOT what I implied. I am not good with computers so i cannot draw you a picture of what I meant, BUT yooure drawing as you made it . Looking direstly at it drilling a hole (and reaming of course for the ground pin) you drill to bissect the small pin with the flywheel or crank half. This creates a Half Moon in each rotationable pressable section (HA hows that for engineering words ) then when you press the pin in place and peen over the HOLES (drilled through crank halves is also OK but not required, Up to youre preferance. peen each end or TIG weld a little blob, Some guys TIG weld the Crank halves and rod Pin. BUT heat and such is better avoided in my opinion.  Just pin the moveable pieces to hold them . They will NOT move Guaranteed and unit is still removable. If you drill Like you drew in youre sketch that weakens everything. Not a Viable Option . Hope this helps., Mike

[RichardShute]

What you are saying is that the pin is axial and ends up acting like a key in a key-way. Albeit, made and fitted in situ.

It's a not uncommon way to do it and, as you say, far stronger than a cross pin.

Richard