Knife Edges for Crank Balancing

[T. Cotten]

I did not disagree with the 'stationary' statement, Folks...

My 'hold and release' technique achieves the same thing.

What bugs me is the drawing. Showing it level would make too much sense.

And the countermass must have bubbles.

....Cotten
PS: Tried to watch the U-tube, but it just sits and spins for me.
Is that George Yarocki?

[TechNoir]

Hello Folks,

A pair of "straight" edges turned up yesterday but I need some advice. They are 750mm, say 2' 6", long so I would need to shorten them to my intended 2' long knife edges but before I use them I want to make sure that they are straight. I don't have a surface plate (it is something that I want/need to get but my small workshop is already at greater than 100% capacity so I have been holding off on getting one) so the obvious thing to do was to compare them against each other. Resting the two beveled edges together (see picture) there is a gap in the middle. Using feeler gauges as a guide I can get a 0.002" gauge between them for about 6" to 8" but not a 0.003". The gap then tapers out over the next 6" or so at both sides.

Now I am aware that this is not definitive, they could be much worse than this if one is convex and the other concave but I don't have a known straight edge to hand to compare at the moment.

So my question is, assuming I can verify how straight these actually are, what is an acceptable level of straightness?

Also, when placing a, say, 20 or 30 pound crankshaft assembly on knife edges then how much flex do you get in them anyway? I can return these to the vendor and swap them for higher quality/more expensive items but before I do I want to gauge what level of straightness I am aiming for?

Thanks






John

[BoschZEV]

I don't have a surface plate (it is something that I want/need to get but my small workshop is already at greater than 100% capacity so I have been holding off on getting one)

The problem with surface plates is they weigh so much that you can't just put them out of the way when they're not being used. Since Nature abhors vacuums and uncluttered spaces they will fill up with items "temporarily" placed on them. But, I digress. With a surface plate determining how straight your straight edges are is, ahem, straightforward. Without it the least expensive alternative I can think of is a precision straight edge, and those aren't all that cheap.

what is an acceptable level of straightness?

That can be calculated. To arrive at a good estimate, assume you have a straight edge that tilt downward by 0.002" over a distance of 1 ft. This makes it an "inclined plane," for which equations aren't needed because we can cheat using an on-line calculator such as the one here:

https://www.engineeringtoolbox.com/i...es-d_1305.html

To see where this leads, assuming there is no friction, and plugging 10 kg (22 lbs) into the calculator along with an angle of arctan(0.002"/12") = 0.01 deg., a force of 0.0017 Newtons is required to pull that weight up the plane. The same 0.0017 N is trying to get it to slide/roll down that plane. Plugging in the acceleration of gravity, this is 0.17 grams. That is, an imbalance of only 0.17 g will cause your flywheel to roll down that slope. But, put another way, if you have 0.18 g on the other side of the flywheel, it will roll up the slope. What this means is if your parallels are bent or tilted by this amount you wouldn't be able to balance your flywheels better than a negligible 0.2 g because you wouldn't know if they were rolling because of weight imbalance or because of tilt of the straight edges. Assuming I haven't made a mistake in my calculation...

Also, when placing a, say, 20 or 30 pound crankshaft assembly on knife edges then how much flex do you get in them anyway?

This also can be calculated. Or, we can again cheat by using an on-line beam calculator. However, I'll leave that as an exercise for the reader because you'll have to enter the dimensions of your straight edge and look up Young's modulus for steel to get a result. Once such calculator is at:

http://www.engineeringcalculator.net...alculator.html

[TechNoir]

Hi Bosch, thanks for taking the time to reply, I know that you are very busy at the moment so your words are very much appreciated.

Since posting my last post I have had a think and a chat to a friend. My lathe bed is a vee way and so whilst it is an old lathe with relatively worn ways the portion of the front way just behind the vee is relatively unworn as the saddle does not touch it. Using this as a comparator I think most of the problem is with one of the straight edges rather than a combination of both.. My friend has a lathe which has had relatively little use since its beds were professionally reground so I am going to compare the straight edges against his lathe bed also. I shall see how they all compare and if the results are consistently the same then I should be able to determine exactly where the high spots on the straight edges are. Then I will decide if I am going to return them or address the high spots myself.

I still might invest in a precision straight edge as per your suggestion, I am sure I will get some use from it more than just checking these knife edges.

Thanks very much for the calculation, Your conclusion:

What this means is if your parallels are bent or tilted by this amount you wouldn't be able to balance your flywheels better than a negligible 0.2 g because you wouldn't know if they were rolling because of weight imbalance or because of tilt of the straight edges. Assuming I haven't made a mistake in my calculation

Seems to say that 0.2g is an acceptable tolerance when balancing these old engines. The balance I have ordered is accurate to within plus or minus 0.1g so if I could reduce the error in the straight edges to, say, 0.001"" then I think we are good to go.

You were also kind enough to point me in the direction of calculating the sag in the "beams"

This also can be calculated. Or, we can again cheat by using an on-line beam calculator. However, I'll leave that as an exercise for the reader because you'll have to enter the dimensions of your straight edge and look up Young's modulus for steel to get a result. Once such calculator is at:

http://www.engineeringcalculator.net...alculator.html

This I had thought about a bit more (maybe because I work in Civil Engineering and there is someone talking about this sort of stuff everyday in the office even if I am not a Civil Engineer myself). I already had plans to beef up the straight edges by sandwiching them between some larger section steel bars. However your first calculation, giving an indication of the effect of slope on the balancing operation, will enable me to determine just how much sag is acceptable and how beefy the sandwich needs to be or maybe just shorten the beam a bit.

I will post an update once I have advanced things a bit further although it might be a week or two due to some seasonal distractions.

Thanks very much once again your thoughts and comments are very much appreciated.

John

[T. Cotten]

...So my question is, assuming I can verify how straight these actually are, what is an acceptable level of straightness?

Also, when placing a, say, 20 or 30 pound crankshaft assembly on knife edges then how much flex do you get in them anyway? I can return these to the vendor and swap them for higher quality/more expensive items but before I do I want to gauge what level of straightness I am aiming for?

John!

I had the edge of my planer blades ground flat as a pair to assure they were matched and "straight"
And yes, a flywheel assembly definitely can make rails sag. I have to place the assembly quite close to supports, even with 1 1/16" rails, and would suggest that the supports be no more than six inches apart.

....Cotten

[TechNoir]

and would suggest that the supports be no more than six inches apart.
....Cotten

Hi Cotten,

Thanks for your input. I had planned supports further apart but given the weight of these old crankshaft assemblies I can see why you recommend 6". I will weigh my 20F and my Matchless and see what level of sag my proposed apparatus calculates as with these weights and adjust the length to suit.

I have 2 straight edges, each 2' 6" long. If my "good" one turns out to be good then if I were to half it I would still have two good edges each 15" long. This would seem more than adequate given your 6" recommendation.

Many thanks to you too for your help with this, I am learning a lot.

John

[T. Cotten]

I learn too, John!

My apparatus taught me that planer blades were not a great idea, although I use smaller ones on a wheelhub, rim, and tire poise.
Somewhere I posted that if I were to do it all over, I would use ground round rod, and now I must add shorter length.

There's complications in holding the blades, as even setscrew tension warped them in the fixture, and can only hold them 'finger tight'.
(Please note the mirror behind my edges in my previous attachment. It is valuable especially when placing the flywheel assembly squarely on the rails, as it is as if you stood back a few feet for perspective.)

An ideal fixture would be monolithic,.. cut, ground, and polished from granite, with perhaps agate 'edges'.

(*sigh*)

Cotten

[BoschZEV]

Seems to say that 0.2g is an acceptable tolerance when balancing these old engines.

I didn't mean to imply that. I actually think you can get by with less sensitivity than that, i.e. 1 g or even worse. More serious issues might be sag and/or the tendency of the relatively wide and thin bars to twist under the weight.

[TechNoir]

I would use ground round rod, and now I must add shorter length.
Cotten

I did take note of thhose comments and my planned setup would not be difficult to adapt to rods or to shortening the length

(Please note the mirror behind my edges in my previous attachment. It is valuable especially when placing the flywheel assembly squarely on the rails, as it is as if you stood back a few feet for perspective.)

Cotten

I saw the mirror and guessed that was its purpose.

An ideal fixture would be monolithic,.. cut, ground, and polished from granite, with perhaps agate 'edges'.

(*sigh*)

Cotten

Oh for an ideal world, but thats a whole other conversation!

I didn't mean to imply that. I actually think you can get by with less sensitivity than that, i.e. 1 g or even worse. More serious issues might be sag and/or the tendency of the relatively wide and thin bars to twist under the weight.

I have pondered the possible distortion of the knife edges and have some ideas for variations on the basic principles. At the moment I am planning on sandwiching the straight edges between flat bar however the outside of the sandwich could be substituted for angle or tee section to increase rigidity.



John

[Peter Cooke]

At the moment I am planning on sandwiching the straight edges between flat bar....

What about milling a slot in a length of ground round bar and clamping your straight edges into the slot. Then, if you wanted, you could rotate 180 degrees and use the ground round bar edge instead. You get extra rigidity and 2 sets of edges to choose from.

[TechNoir]

What about milling a slot in a length of ground round bar and clamping your straight edges into the slot. Then, if you wanted, you could rotate 180 degrees and use the ground round bar edge instead. You get extra rigidity and 2 sets of edges to choose from.

Peter, that is a really good idea. The one drawback (for me) is that I don't have a functional mill at the moment although I am hoping to remedy that in the near future. If I do then I will give it some serious consideration.

Thanks,

John

[Peter Cooke]

No problem John,
I have to admit I've not done it myself. I was reading through the thread and it wasn't until you mentioned sandwiching your straight edges between flat bar that it occurred to me. I'll be needing to balance my VL engine somewhere down the track so am interested in the topic. There are many ways to skin the cat...

[T. Cotten]

My apparatus is flawed, Folks,...

But it illustrates the pitfalls of attempting to clamp to planer blades, as a firmly set set screw screws them.

(Horizontal against the blade in the attachment.)

Ideally, nothing should need to be retained by anything but gravity.
Round stock is quite secure in v-blocks, especially with the weight of the flywheel assembly upon it.
The trick is getting the v-blocks up high, rock solid.

And a steady hand placing the wheels on them!

...Cotten

[TechNoir]

Hi Cotten, thanks for your input again. More food for thought. I now have several ideas for my edges and a few days to ponder my plans.

Thanks to everyone who have contributed to both this thread but also everyone on the AMCA forum for your help and words of advice over the past year.

Seasons greetings to everyone.

John

[TechNoir]

Seasons greetings to everyone, I hope you (for those who celebrate it) had a great Christmas. The last of my guests left earlier today and so I have less distractions although it was nice to have them, it is also nice to have some peace.

Thanks to BoschZEV for taking me on a trip down memory lane, following Bosch's post I had a look at some options for my knife edges. The trip down memory lane was because I did a year of Structural Mechanics in my first year of University in 1996.

Because I have a few different ideas I put it into Excel so I could compare the various amounts of sag for the different knife edges I have in my head. The first thing I noticed was that one of my options gave an angle of slope similar to the theoretical example that Bosch posted about.

.....plugging 10 kg (22 lbs) into the calculator along with an angle of arctan(0.002"/12") = 0.01 deg., a force of 0.0017 Newtons is required to pull that weight up the plane. The same 0.0017 N is trying to get it to slide/roll down that plane. Plugging in the acceleration of gravity, this is 0.17 grams. That is, an imbalance of only 0.17 g will cause your flywheel to roll down that slope.

However my maths gave a result for the mass of the imbalance needed to move the 10kg flywheel to be a factor of 10 greater than the example. I double checked the on-line calculator that was posted and I think Bosch made a typo in his original example. I think it should be 1.7g rather than 0.17g. What do you think?

Anyway, my plans for the knife edges have settled on 300mm between supports and using the beveled straight edge with a stiffening bar, probably the thicker of the two in my table below. (note the straight edges are approx max 38mm x 5mm with the bevel starting 8mm from the edge) In order to get the maximum stiffness in the edges I will need to fix the straight edges to the bars. I plan to use something like Loctite Epoxy Adhesive rather than mechanical fixings although I may mill a small "shelf" into the support bar so that the load is also physically bearing onto the bar rather than being transferred to the bar 100% by the adhesive in shear.

I only have bathroom scales for weighing the crank assemblies but I am sure they are accurate enough for this estimation. My 20F crankshaft assembly and 2 rods weigh 12kg's and a 1953 Matchless single crankshaft and rod came in at 12.6kg's. So I based my calculations on a weight of 13kg's.

The results (assuming I have not made some stupid error) are summarised as follows:




If I were to go with round bar for the edges then to get the sag minimised I would need quite large diameter bars. I would then wonder if the width of the bar would cause issues when trying to fit the crank on them?

The final decision is a couple of weeks away at least because I have some, essential, distractions to take care of. I have a small mill that I am rebuilding, it is not far away from being functional and I want to get it to that stage before I look at these knife edges.

Also, I bought a new to me Lathe (an upgrade from my current one) which is being delivered next week. It will need some attention before I can use it because it is 3 phase so I need to wire it to retain it as 3 phase but power it through a vfd. Also, remember my workshop is at greater than 100% capacity, in parallel I need to sell my current lathe.

John