Another Rupnow 1 by 1

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I have an "impatient" approach. Make it go, put it on show, talk about it with those folk who wish they could make "beautiful " models, but too shy to try. More join the Club that way. And I respect the perfectionists, just don't try and sing in the same choir, even if I do read the same hymn sheets.
I'm human and fallible, and not too shy to show it. All my models end up working, but I can't say any are pretty, at the show standard of stuff displayed here. That's (my) life!
K2
 
I never thought of dog washing. My fall back on "I'm going to start a less demanding hobby" was always stamp collecting. There are only two steps:
Buy stamp.
Paste in book.

Neither can be done to even .001"

Dog washing sounds easier, although probably messier. Probably depends on the dog.
Just so you know, I am a stamp collector as well as machining, welder, all round King of Fools (KoF) plus a foo other things too.
 
Hi CFLBob!
I looked at this part for hours yesterday. Yes, it might well actually be usable. Yes, it's going into an engine where it will never be seen so "ugly is OK." I just know I can do better, so I'm going to do that.

Maybe, just grind and it will be beautiful again
If you do not like , make a new one
Personal: I like every part of the engine to be beautiful
But often I can't do it, just accept it :D
 
Second attempt about to get underway. I need to call the center of that left hand screw X=0.500 and Y=1.000 to use files I've already created. I think it's currently those coordinates, but I'll triple check that.

Take2.jpg


I'm going to rough around the part with a bigger end mill and a contour cut farther offset from the outline to leave a larger margin around the part. Not sure what - maybe 1/8" instead of 1/32. I'd like to eliminate cutting slots but that's easier at the wide end than at the narrow part of the rod.
 
Go Bob Go......

One thing I do to insure the hold down screws are in the exact center, is to drill the holes with the CNC mill set to 0,0, then tap them in the vice. It is a little akward, but insures the center of the conrod holes are exactly where your CNC program thinks they are in both the X and Y axis.


I have a good feeling about this one. :)
 
Go Bob Go......

One thing I do to insure the hold down screws are in the exact center, is to drill the holes with the CNC mill set to 0,0, then tap them in the vice. It is a little akward, but insures the center of the conrod holes are exactly where your CNC program thinks they are in both the X and Y axis.


I have a good feeling about this one. :)

The first operation went as smooth as can be.

RoughPass-375EM-125-margin.jpg


This was done with a 3/8 dia. carbide, 4-flute EM. My intent was to make it big and wide enough to get rid of all of the rest of the rough - and it did. I did the first two passes with compressed air coolant and cutting oil in the slot, then added the Fogbuster mist for the last two passes.

This left a 1/8" margin over the part design. I only checked one detail, the diameter of the small boss on the left and got 0.252 difference between CAD design and metal - 0.126 on each side.

My plan was to do the circular bosses and thin out the rod between the two big areas, but while my G-code checker (GWizard Editor from CNCCookbook) told me the code was fine, Mach3 barfed at it. In the worst way; the window that tells you what the problem is was too short for the full message. The code is the way I did two semicircles separated by 0.050.

The bulk of the code was the way I did the bosses on my Webster's conn rod.
 
Last edited:
I found the silly error in my Gcode for the circular bosses. I left off the negative sign on J= number on the circles that needed it.

G91.1
G01 X0.907 Y1.000 Z0.125 F25
G01 Z-0.031 F6.5
G03 X0.907 I-0.406 F12
G01 X2.437 Y 1.00
G01 X2.437 Y0.502
G01 X2.937
G02 X2.937 Y1.503 I0.00 J0.500
G01 X2.987
G02 X2.987 Y0.503 I0.00 J-0.500
G01 X2.937

After verifying the outlines still seemed correct in my GWizard Editor and then air cutting above the part, it was on to the real deal.

Bosses_Cut.jpg


Now I have a puzzle. The bosses need to be cut on the far side, too, and the body of the part is still oversized by 1/8" in all directions (1/4" overall).

It seems to me I should cut the body to final size first, then flip it over and re-run this code to cut the bosses on that side. The only drawback is that the final contour cut requires me to change cutters, which may require me to change the holder. This file used a 3/8" EM, while the fine contour needs a 1/4" EM - the inside radius where the boss on the left meets the rod is 1/8".

If I flip it and then cut the bosses, I use this EM and then change to the smaller one.

It's two tool changes vs. one. Not a big deal, since we're not planning production of a million of these and every second counts.

Any inputs? Anybody see anything I'm missing?
 
The machining of the part proceeds with the milling done. It will be a precise drill press for the next operation and then drive a slitting saw blade.

Side2_Done.jpg


To answer my own last question, I did the final contour first in the position shown on the previous post - that required a tool change from a 3/8 EM to a 1/4" EM. Then I flipped the part, changed back to the 3/8" EM and cut the circular features. After that was finished, I removed the #8 screw (on the left) and then finished that hole, which is reamed to 0.188.

In my mind, this result is much better than the previous approach.

Still to do is to use the CNC mill as a precise drill press. Twice. I need to drill and tap the holes for the screws that hold the cap on. Then I use it to hold the slitting saw so I can cut off the cap with the slitting saw. Finally, I'll attach the cap with the screws I just made holes for, and turn the mill back into a drill press to drill and ream the big end to 0.375.
 
All sounds logical and straight forward. Enjoy!

I am in the middle of making the valve chest innards for the steam pump I am making. Lots of milling and drilling... The secondary valve works with air, now making a the primary... must try and do some photos...
K2
 
Bob,

It is looking really good. What is your concern with changing end mills? The X and Y should remain dead on and the Z height needs to be changed to account for any difference in height between the two end mills. I will machine with the first size end mill, then bring the end mill down and touch off on the top of the vise, record the Z value displayed on the DRO. Load the second end mill, again drop down and touch off Z at the same spot on the vise, set the Z value to the one recorded. Then machine with the next tool path.

Greg
 
Bob,

It is looking really good. What is your concern with changing end mills? The X and Y should remain dead on and the Z height needs to be changed to account for any difference in height between the two end mills. I will machine with the first size end mill, then bring the end mill down and touch off on the top of the vise, record the Z value displayed on the DRO. Load the second end mill, again drop down and touch off Z at the same spot on the vise, set the Z value to the one recorded. Then machine with the next tool path.

Greg

Greg,

That's what I do. Basically, any time I change the setup, I get concerned. Work holding is everything. I haven't adjusted the Y axis setting in the mill since I started working on version 1. A full month? I have reset X with an edge finder when I turned the first one over, and then again with this one. Tedious but I trust it to under 1 thou.

Yesterday, after I cut the circular areas and I switched to the 1/4" dia. EM, I made a mistake that I spotted soon enough to not damage anything. I went to the flat area between the raised bosses and set the Z to 0.000. Only it's not. It's -0.062. I noticed doing the fine contour pass (taking off 1/8" all the way around the part) that the heights looked wrong. That's when it hit me what I did. Stopped Mach3, reset that zero to what it should be, stripped the last pass around the part out as a file and ran the last pass by itself.

It wasn't a big deal, it just would have cut the tooling piece deeper than I want.


Bob
 
Greg,

That's what I do. Basically, any time I change the setup, I get concerned. Work holding is everything. I haven't adjusted the Y axis setting in the mill since I started working on version 1. A full month? I have reset X with an edge finder when I turned the first one over, and then again with this one. Tedious but I trust it to under 1 thou.

Yesterday, after I cut the circular areas and I switched to the 1/4" dia. EM, I made a mistake that I spotted soon enough to not damage anything. I went to the flat area between the raised bosses and set the Z to 0.000. Only it's not. It's -0.062. I noticed doing the fine contour pass (taking off 1/8" all the way around the part) that the heights looked wrong. That's when it hit me what I did. Stopped Mach3, reset that zero to what it should be, stripped the last pass around the part out as a file and ran the last pass by itself.

It wasn't a big deal, it just would have cut the tooling piece deeper than I want.


Bob

Yep, I have done the same thing (actually ruined a part that way), that is why I touch off the vise, write the z height down, instead of touching off on the part again. Because I might have machined off the top!

But watch those minus signs, I set a Z height one time and forgot the minus sign, and the Z axis took a rapid move dive into my part. Scared the bejesus out of me.

You are doing all the right things.
 
Hi Bob, I like what you are doing!
My "memory" (Datum zero) is the chalk mark on the rotating scale... (NON-"Electronic Digital" living). Works every time for me... Just slacken the lock screw, rotate the scale and set at the chalk-marked division. May be of use to those with MANUAL tools? I agree with deciding whether to "zero" from part or machine datum. I always use "part" to avoid moving the part from its alignment with the previous tool axis (Worn feed-screws mean the settings and measurements are only repeatable from one direction anyway).
A point to note. NO machine is perfect. So correctly dimensioned parts won't be "random" dimensions strewn all over the drawing (Draughtsmans' normal practice), but ALL set from the appropriate datum (rarely seen on "Amateur" plans and especially CAD!). The process should always be (Or so I was taught in the days of worn WW2 machines!):
  • Set the part to machine the part datum (Face and location sides or centre/hole).
  • PLUS any Datum precise machinings required before changing settings. (Mark datum faces, and annotate the drawing!).
  • Plus any "non-datum " based shapes accessible in this setting.
  • "Turn over" to set ON the datum face, and up to datums that were machined in the first setting.
  • Machine all the previously inaccessible bits to the datums. Then finish by machining all the other non-datum specific shapes.
Remember that the designer didn't think like a machinist, so don't always blindly follow his dimensions. (That was one of the most important rules I learned as a designer!). When making all my jobs I find I have to re-calculate some dimensions to DATUM. And if you don't do it right before you start machining, you'll end up making lots of "trapezium lozenges"! (Speaks a voice of experience). And my miller is a known fraction of a degree off true... in each of the dimensions (degrees of freedom)! But by working to datums, the finished job always works OK. Remember the OLD rule, "right first time every time!" - and Irish Quality Assurance (Based on the Navy's "CHECK, CHECK, CHECK!"): "To be sure, to be sure!".
Now Please teach me where I am wrong, because I have forgotten stuff, and don't want to write something that the "less knowledgeable" will take as gospel - and then get it wrong!
(Now I must go and "do as I said", not "what I did"...! - I am only human, after all.).
Cheers!
K2
 
My problem is that since I have no actual training at all in how to do any of this means I pretty much only learn from my mistakes.

To that point, if my drawings don't show a datum point, how do I pick one? Come to think of it, I might have done that on this part. As I worked on the drawing after importing it from Brian's .pdf, the small screw hole (in the left boss) ended up centered on (0.500, 1.000) so all the tool paths I had generated were based on that coordinate system. To use those files, I had to fixture the part with the hole there.

What I did when I made this version 2 of the part was use the center of the small hole as my reference for X and Y. I have a center finder that has a point one end, and that seemed best for using in a hole because tiny amounts of being off center seem easy to see.
 
As of this evening, the conn rod is finally done.

ConnRodDone.jpg


I had transferred the rod off that fixture to drill the holes for the cap screws on the big end and cut the cap end off. Then I had to re-transfer it back onto that fixture to dill and ream the big hole on the right. I think I reset zeroes three times today.

One little gotcha is that the plans call out #5-40 screws. I would have sworn that I bought those but searching the shop turned up nothing. I don't a drawer in my box for #5 and they weren't with the other screws I bought in preparation, so either I put them "someplace safe" (really safe) and they'll turn up now that I'm not looking for them or I never ordered them. I thought about just using 4-40 but went up to 6-32. Still fits and may even be a touch stronger.

In this view you can easily see how much better this came out than version 1.

ConnRodDone-2.jpg


Another part to go in the "Done" box. It's almost a month ago that I got started on version 1 of this, which led to my CAM discussion. I don't know how many parts are left to do, and I don't really want to know. If I make one part per month I'll be working on this engine for years. When I started the cylinder (in December!) I figured the next parts would be the matching piston because it gets sized to the cylinder, which means I need this connecting rod and the wrist pin. Those are lathe parts so the adventures in CNC milling are over for the moment.
 
Keep going Bob, I will chuck-in ideas as I think of them, but you have made what seems to be a very good con-rod!
- Just odd ideas (from what I remember from when I worked on Engine design years ago!):
The con-rod-bolts. What material are they? They "look OK", but as an engineer, that is where I always see a big "CAUTION" sign! The tensile strength of the con-rod bolts are severely stressed, and fatigued, so need to be a VERY robust design (usually) to cope with the stresses and resist fatigue. Typically, the endurance limit is less than half the tensile strength of a bolt (the single cycle failure point). EVERY corner of the bolts is usually radiused with quite a large radius, unlike simple "bought" bolts. Never having design the bolts, I cannot advise much, but on car engines it is common for the bolts to be 120ton steel, often forgings, heat treated, and very complex shapes. At 2000rpm (a pretty slow model?) the rods perform 120,000 cycles in 1 hour of running, or 1 million cycles in little more than 8 hours running. Of course a car engine/industrial engine can be expected to perform thousands of hours running, and typically up to 6000rpm or more. (Motorcycle engine s easily double that). So lifetimes are designed for hundreds of millions of cycles.... for reliability. Can you get some high tensile hex. socket cap screws with solid shanks through the most part of the rod? (I can only see the heads, not shanks, of your bolts) - preferably installed in "sized" holes (reamed?). (I guess you have done so?). And another idea... the washers under the bolt heads are normally precision ground so they do not create a skew face for the bolt heads to sit upon. The stress raiser of even the tiniest "out-of-square" between the axis of the bolt and the face under the bolt-head can dramatically reduce the life of con-rod bolts, and is OFTEN a cause of failure of the bolt-heads. (Ask guys that race engines!). Many engine makers specify replacing big-end bolts with brand-new parts EVERY time that the rod-end is dismantled, and re-assembled. It is that critical. Models mostly survive because they have very short running periods, at slow speeds, and low loads. Except aero-engines.
I guess there will be some simple maths somewhere on the web to determine bolt sizes against your design? I'll have a look when I get some time. (DIY at the Daughter's house today!). AND often the threads are special rolled threads, as machined threads are simply too prone to early life failure at the "pre-engineered cracks" in machined thread roots. Nuts also need to be as good as the bolts.
Bolted Joint_R1.doc (live.com) - may be of interest/relevance?
Also, based on root diameter of thread, the #6 x 40 is about 13% "stronger" in tension than the comparable #5 x 40. (88% of tensile stress), because of the increased thread root area... Bolts threads should be as "perfect" as you can get them, lubricated and torqued accurately to the pre-load so they are never slack in the dynamic loading on the rod. (allowing of the coldest start-up - differential expansion). Best Way to Improve Fatigue Resistance of a Bolted Joint - Nord-Lock Group may be of interest?
Well done so far with the rod! I'm sure your expertise (and CAD-CAM) has set datums that work, (as otherwise you would have a lot of scrap!).
Most interesting from my chair.
Thanks for posting your process and thinking for making the rod.
K2
 
Last edited:
Keep going Bob, I will chuck-in ideas as I think of them, but you have made what seems to be a very good con-rod!
- Just odd ideas (from what I remember from when I worked on Engine design years ago!):
The con-rod-bolts. What material are they? They "look OK", but as an engineer, that is where I always see a big "CAUTION" sign! The tensile strength of the con-rod bolts are severely stressed, and fatigued, so need to be a VERY robust design (usually) to cope with the stresses and resist fatigue. Typically, the endurance limit is less than half the tensile strength of a bolt (the single cycle failure point). EVERY corner of the bolts is usually radiused with quite a large radius, unlike simple "bought" bolts. Never having design the bolts, I cannot advise much, but on car engines it is common for the bolts to be 120ton steel, often forgings, heat treated, and very complex shapes. At 2000rpm (a pretty slow model?) the rods perform 120,000 cycles in 1 hour of running, or 1 million cycles in little more than 8 hours running. Of course a car engine/industrial engine can be expected to perform thousands of hours running, and typically up to 6000rpm or more. (Motorcycle engine s easily double that). So lifetimes are designed for hundreds of millions of cycles.... for reliability. Can you get some high tensile hex. socket cap screws with solid shanks through the most part of the rod? (I can only see the heads, not shanks, of your bolts) - preferably installed in "sized" holes (reamed?). (I guess you have done so?). And another idea... the washers under the bolt heads are normally precision ground so they do not create a skew face for the bolt heads to sit upon. The stress raiser of even the tiniest "out-of-square" between the axis of the bolt and the face under the bolt-head can dramatically reduce the life of con-rod bolts, and is OFTEN a cause of failure of the bolt-heads. (Ask guys that race engines!). Many engine makers specify replacing big-end bolts with brand-new parts EVERY time that the rod-end is dismantled, and re-assembled. It is that critical. Models mostly survive because they have very short running periods, at slow speeds, and low loads. Except aero-engines.
I guess there will be some simple maths somewhere on the web to determine bolt sizes against your design? I'll have a look when I get some time. (DIY at the Daughter's house today!). AND often the threads are special rolled threads, as machined threads are simply too prone to early life failure at the "pre-engineered cracks" in machined thread roots. Nuts also need to be as good as the bolts.
Bolted Joint_R1.doc (live.com) - may be of interest/relevance?
Also, based on root diameter of thread, the #6 x 40 is about 13% "stronger" in tension than the comparable #5 x 40. (88% of tensile stress), because of the increased thread root area... Bolts threads should be as "perfect" as you can get them, lubricated and torqued accurately to the pre-load so they are never slack in the dynamic loading on the rod. (allowing of the coldest start-up - differential expansion). Best Way to Improve Fatigue Resistance of a Bolted Joint - Nord-Lock Group may be of interest?
Well done so far with the rod! I'm sure your expertise (and CAD-CAM) has set datums that work, (as otherwise you would have a lot of scrap!).
Most interesting from my chair.
Thanks for posting your process and thinking for making the rod.
K2

They're the best kind of screws: chosen for size from from a junk box full of random screws!

More seriously, they seem to be stainless, but are likely to be made for a different purpose than holding a conn rod together. Probably 18-8 or 304 stainless. The extent of the information I have on what to use is one note: "D & T FOR #5-40 SHCS". These don't have a longer shoulder, and are threaded all the way to the cap (or around 1 thread before the cap).

The introduction of the reality for the screws in this application for high performance engines is interesting engineering. I think that a 1" bore, 1" stroke engine (pi/4 or 0.785 cubic inches) that will run for a few minutes in its life and get put on a shelf isn't likely to ruin them.
 
Back
Top