how to locate each holes positions relative to the center of engine shaft?

[blane.c]

You can buy a 5C spin index cheap. CDCO Machinery Corp. (cdcotools.com) 5c collets cheap too. CDCO Machinery Corp. (cdcotools.com) Just make a jig to hold index and either fit a collet or machine a collet to fit the crank, you can then dial in at 1 degree increments. Finer than that make the jig holding the index moveable at the set increments necessary.

Tried to make it click to appropriate but just goes to home page CDCO, so in search box at left look for 5C spin index, or 5C collet. They have step collets to 4" dia. and 5c collet chucks too.

 

[proppastie]

If you measure to pins

In theory absolutely correct however one needs accurate tools and the skills of a class A machinist or tool maker to match parts with press fit dowel pins and bearings.

 

[PMD]

In theory absolutely correct however one needs accurate tools and the skills of a class A machinist or tool maker to match parts with press fit dowel pins and bearings.

I think I forgot the part about: once you have an x-y file use a digital readout or full CNC milling machine to drill and ream the holes. I have actually done these things with manual layout in a drill press, but as you surmise, it is really difficult to get the tolerances down to something that works with the dowel pins. thanks for catching that for me/us.

Using a spin index for layout of such a large part adds potential for significant angular error and does little to provide for getting the radial distance correct....thus just use 2 axis mill.

 

[proppastie]

Useing a doughnut/guibo on the flywheel

I had to look that up....using one of those is way over my head.....but I would think a belt would be able accommodate .006" out of round of "theory" "toroidal motion" (which I have never seen).If you have .006 side to side movement in a crankshaft.......you have problems which need to be addressed with new bearings....... in my opinion.

 

[dog]

I had to look that up....using one of those is way over my head.....but I would think a belt would be able accommodate .006" out of round of "theory" "toroidal motion" (which I have never seen).If you have .006 side to side movement in a crankshaft.......you have problems which need to be addressed with new bearings....... in my opinion.

The figure of .006 play was referencing a dry crank with an extension,and yes is theoretical.
The point (well taken) is that useing the crank
center to locate the pins and bolt holes is prone to errors.
An adjustable arm bolted to the crank flange that could hold 2"pins of the size
the matches the holes in the block could be used in conjunction with a dial indicator on an
already checked rotating asembly,then perhaps
good numbers could be aquired in a home shop,and sent out for cnc made parts,or a laser cut template to verify the fit first.
In my case,the local shops can use any file I can bring them,but they charge more to make
up a file for a one of than the part costs.

 

[PMD]

The figure of .006 play was referencing a dry crank with an extension,and yes is theoretical.
The point (well taken) is that useing the crank
center to locate the pins and bolt holes is prone to errors.
An adjustable arm bolted to the crank flange that could hold 2"pins of the size
the matches the holes in the block could be used in conjunction with a dial indicator on an
already checked rotating asembly,then perhaps
good numbers could be aquired in a home shop,and sent out for cnc made parts,or a laser cut template to verify the fit first.
In my case,the local shops can use any file I can bring them,but they charge more to make
up a file for a one of than the part costs.

Let me know when you try that how close you think you can get angular displacement!

 

[Hot Wings]

<, >> you end up with a set of numbers that can easily be turned with basic trig << >>

And in CAD, like SW, no trig knowledge needed. Just draw a bunch of circles thru the points as needed, set them =, or the points common to 2 or more circles as a starting guess and let the software do the trig. Some assumptions need be made but the original designers were generally pretty lazy (I know I am) and tend use common centers, nice even angles and "simple" numbers for the dimensions.

 

[proppastie]

a dry crank

the bearing clearance with new plain bearings or ball bearings on a crank shaft is so small that you should not have much out of round. If you look at new limits for Lycoming bearing clearance or crank out of round you will see what I mean. Modern automotive engines I believe are even closer tolerance because they are liquid cooled, Modern NC machines will position, turn or bore to less than .0005 .......and dry or under pressure has nothing to do with it as far as I am concerned. As I said before you will need the skills of a Class A Machinist if you expect to do this .......if you do not have those skills you are wasting your time.

 

[PMD]

And in CAD, like SW, no trig knowledge needed. Just draw a bunch of circles thru the points as needed, set them =, or the points common to 2 or more circles as a starting guess and let the software do the trig. Some assumptions need be made but the original designers were generally pretty lazy (I know I am) and tend use common centers, nice even angles and "simple" numbers for the dimensions.

And THAT is why in this world of digitalis so many people can make their mistakes with unerring accuracy. In flight, we might site situational awareness - in fabricating it is called being sloppy and lazy.

Your observation about probability of some sort of order and consistency in locating/dimensioning is often spot on, but THAT is why you need to do the work - because things are NOT always so.

 

[dog]

the bearing clearance with new plain bearings or ball bearings on a crank shaft is so small that you should not have much out of round. If you look at new limits for Lycoming bearing clearance or crank out of round you will see what I mean. Modern automotive engines I believe are even closer tolerance because they are liquid cooled, Modern NC machines will position, turn or bore to less than .0005 .......and dry or under pressure has nothing to do with it as far as I am concerned. As I said before you will need the skills of a Class A Machinist if you expect to do this .......if you do not have those skills you are wasting your time.

I aint gona machine nuthin,just design and
make a pattern/file. For Class A+ machinists to follow.Get a test template made, verify the pattern.Email the file,and the part will just be waiting when I get home one day.

 

[proppastie]

You will need to send it out to be measured too, because it will take a skilled person to do that or a not so skilled with some expensive equipment.

jig bore services - Google Search

 

[Hot Wings]

because things are NOT always so.

And getting to be even less likely with items designed using CAD/CAM. The old days of designing so the guy in the pattern shop has an easy day are pretty much gone.

 

[proppastie]

Because the pattern shop has been replaced by 3D printed patterns?

 

[Hot Wings]

Or skip the pattern all together and just let the machine make the part.

For the typical HBA guy reverse engineering a part in the workshop out back we use what tools we have on hand as cleverly as we can. Results vary..........

 

[dog]

Let me know when you try that how close you think you can get angular displacement!

How about an abitrary reference point ,three feet from my center,find my bolt hole,extend
the line out to the same diameter as my reference,measure, compare to the same
radius as my bolt, repeat as many times as nesessary to get a grouping,throw out the outliers and average the rest.
Its slow,but it works.

 

[proppastie]

How about

ok how about this.....look up "bullet nose dowel pin" and one dowel is round the other (should have 2 dowel pins for fixed location of the bell housing or your plate) you will be able to not need as quite an exact location of your dowel pins as along the CL along the two dowels some float is tolerated yet 90 deg. to that line the plate location will be fixed......then with clearance holes for the bolts perhaps you will be able to mount and hold location of your plate with less than perfect measurements of your methods........if the crank center is not fixed to your plate because you did not add a bearing to your small pulley the exact location of your center of the crank is not critical because your tensioner will adjust your belt....that being said when you are all finished if you have not calculated or allowed for TV (torsional vibration) your reduction may not work.....and I can not help you with that because I do not know how to do those calculations.

 

[pfarber]

Would the answer be to simple ream and press your OWN dowel pins? You could have a class 3 fit with bolts and then drill/ream/press your own dowels. As many as you like.

As for concentricity with the shaft, well, hows about leaving the center bore .005 -.005 + 0.000 under and then put a flycutter or reamer or something on the shaft and let it self center?

 

[reo12]

Locating holes relative to the crank centerline. Dimensionally drawing? Is it for manufacturing or is it for determining what dimensions the part was made to? Using precision tooling? If with precision tooling - what kind of precision tooling is available? What level of precision is required? Remember - these are multi - dimensional measurements. There are more dimensions than simply x amount from crank center and x number of holes - and if such holes are equidistant about the centerline. If not - that needs to be described. What amount of error in location and linearity is allowed?

Improper dimensioning and open choice of machine operations could lead to someone simply locating the hole array using mensuration. Then center punching the locations and drilling the holes on a drill press that has a tipped table. That may function for a sheet metal part that has 3/32" over size holes but a 2" thick part with tight fitting bolt or dowel pin holes is an entirely different matter.

 

[bmcj]

Not a solution for a precision final cut, but can give you a good first pass estimate… I have used a telephoto lens from a long distance to get a reasonably accurate flat view. The longer the distance from the camera, the less the parallax error.

 

[Tiger Tim]

Couldn’t this just be done with a bare engine block, a plate (probably metal but plastic or wood or MDF or a pizza box could be made to work), and a set of transfer punches? That will reduce it down to either take measurements in two dimensions or just be a pattern for whatever you do make.