machines which I have made, am making, or intend to make, and some other stuff. If you find this site interesting, please leave a comment. I read every comment and respond to most. n.b. There is a list of my first 800 posts in my post of 17 June 2021, titled "800 Posts"

Tag: machining mistakes

Learning from MISTAKES

It is said that you learn from your mistakes.

Actually, I think that you TRY to learn from your mistakes.  It makes you feel a bit less stupid when you make mistakes.  Or, at least, it puts a bit of worth into having made a mistake.

At the rate of my mistakes in recent days, I should be turning into an Einstein.  Somehow, I doubt that is happening.  I think that I will stick with my motto of avoiding making the same mistake more than 3 times in a row.

Yesterday, I was cutting gear teeth into the big gears which I had cast a few days earlier.  Due to the mistake of not allowing for shrinkage of the castings, the wheels were almost 1mm smaller diameter than planned. (#1.). That meant that the wheels would have 55 teeth, rather than the planned 56 teeth.  Not a big problem, just an annoyance.

Also, I had not cast a shaft in the gears, or any method of holding the castings for teeth cutting, (#2) so I had to drill a central bore, insert an 8mm shaft, and solder it in position.  I should have used steel for the shaft, but for some reason I chose brass. (#3.)

Then, I could not find my module 1 gear cutters.  The  cutters, I knew, because I had seen them recently, were in a small cardboard box.  I went to where they were usually stored, and they were not there.  OK.  I must have put them somewhere else.  Then spent 2-3 hours searching every shelf, drawer, bag, box, floor, machine…. you get the picture.  And could not find them.  So I searched my vehicle, the other shed, and could not find them.   Had I loaned them to someone else?  Surely I would remember that?  So I went and had a cup of coffee.   Hmm.  What next?  OK, start searching again, going over previously searched spots in case I had missed them.  So, first to the cupboard where they were usually stored.

And there they were.  Exactly where they should have been.  Well bugger me!  (#4).  There was a small upside to this long and frustrating search.  I had found a few other tools which I had not seen for quite a long time.

So then I commenced the gear cutting.  55 teeth into the bronze wheel.  The setting up of the CNC rotary table on the mill, attaching the cutter to the arbor, cutting teeth marks onto the circumference to check my calculation of the number of degrees per tooth (360/55 = 6.545454º….).  All going well.   Each tooth required a cut 2.66mm deep.  But I was a bit concerned about that 8mm brass shaft.  Would it bend under the pressure of the tooth cutting?

So, I cut the teeth in 2 stages.   1.5mm deep for the first cut, 2.66mm for the second.  The first cuts went well.  Looking good.  After a complete circuit, the partly cut teeth were all even, and gleaming.  So, onto the full depth of 2.66.

DISASTER!  (#4.). When I attached the cutter, I had carefully placed it on the arbor so that the securing nut would tighten, rather than loosen with the cutting process.  But I had got it wrong.  Totally wrong, and it loosened!   BAANG.  I hit the big red button.  The mill spindle stopped, but the now loose cutter was still spinning wildly.  What if it comes off while spinning like that?  Would it fly across my workshop, cutting whatever it hits? Like me?

When I bought a woodworking spindle moulder many decades ago, an uncle advised me to NOT use it.  He had seen a co-worker die, when a cutter flew off a spindle moulder, slicing into his abdomen.  I did use the new spindle moulder, but with great caution, and never had any problems.  

Visions of the gear cutter slicing into my belly.

But, the cutter slowed, then stopped, to my relief.

A quick look at the workpiece was disheartening.

The brake drum attached to the gear was bent .  A spoke had a big gash.  The tooth which was the culprit for this disaster was gashed too deep.  and the shaft was bent to a crazy angle.

First thought.  “Bin it and start again”.   Steps to make another big gear.   1.  3D print a new PLA gear (with machining allowance this time, and include an oversize shaft).  2. Make a casting tree. 3. Mix and pour the investment. 4. Dry, burnoff, and bake the investment. 4. Melt the bronze and pour. 5. Machine the gear blank, cut new teeth.  2 days estimated, if all goes well.

Oh well.

No photos of the damaged gear.  Not in the mood.  But I did take a closer look.   I wonder if it can be repaired.  Maybe the distorted brake drum could be bent back into shape?  Or cut off totally, and silver solder on a new one? Fill the defects, and tidy up the damage.

I decided to call it a day.   But first I used a hammer on the damaged bronze gear brake drum.  Somewhat to my surprise, it went back into reasonable shape, without cracking.  So, a bit more tapping, and it looks quite good.  It will require a bit of filling and filing, and machining, but maybe it is resurrectable.

After a sleep, and fresh look, I relented and took a couple of photos.



But what about finishing cutting the gear teeth?  It will need a new shaft.   My plan is to machine off the bent shaft, insert a new steel shaft while holding the external circumference of the gear in the lathe chuck.  Might just work.  If the over-gashed tooth looks too odd when I finish cutting the teeth, I will fill the void with bronze or silver and re-machine it.  Watch this space.


So Many Mistakes! Am I Too Old for this hobby? Or is it the heat?

Having completed the model Trevithick dredger engine, and not having an inspiration to start another major build, I decided to make another steam driven boiler feed pump.

Earlier this year (2019) I made a horizontal, duplex, twin cylinder feed pump for the 6″ vertical boiler, but I had also purchased the castings and plans for a vertical, single cylinder feed pump, not having decided which version to fit to the boiler.  The horizontal twin version fits and functions very well, but I decided to make the vertical version while I am thinking about another major build.

Screen Shot 2019-12-30 at 11.01.18 am.png

This is what it will look like.  Hopefully.  Single acting 3/4″ steam cylinder top, and 1/2″ water pump bottom.


The plans, O rings and castings.  The castings have been cleaned up on the RadiusMaster, and the steam cylinder (top) is almost finished.


The 7 pages of plans are excellent.  Imperial measurements and fasteners.  I will use metric fasteners.


But I work in metric.

So over the past few days, excluding the ones over 38ºc (100ºf), I have been machining the gun metal castings.  And making a real mess of it.

The Mess.

  1. The steam cylinder bore.  Bored with a boring head on the milling machine.  Turned out nicely, but I decided to run a 3/4″ reamer while it was set up on the mill, thinking that the dimension would be more accurate.  I did not notice until too late that the reamer was damaged.  It badly scored the bore.  I considered making a new cylinder from bar stock, but used the boring head to remove the scores.  Now 0.75mm oversize.  Annoying but not fatal.
  2. Steam passage not centered.


    The steam passage in the cylinder cutout is meant to be centered.  It is off at a 15º angle, and is centered with the cylinder top, but not the bottom of the cutout where it should be.   OK, it will not be seen, will not affect the function.  Just a trivial mistake.  That is the final oversize bore.

    3.  This one could have been fatal.  All of the center drill holes for the screws and steam passages in the cylinder valve face were off by about 1mm.  The workpiece had moved in the milling vice between setup and machining.  I really thought that this would probably require a new part, but I decided to proceed and see what eventuated.


    The middle 6 holes are the steam inlet and exhaust passages.  Fortunately they are in the correct vertical position, and have just been widened horizontally by the incorrect centre drill holes, which should not effect the function.  The screw holes merged into the incorrect holes, and were pulled a fraction laterally, but should be OK.   At final assembly I will fill the incorrect holes with something, probably epoxy or gasket goo.

    4. This was the most obvious error.  Moved the mill table in the wrong direction, and the rectangular hole ended up with an extension.  I don’t think that it will effect the function.  And it wont be seen by anyone except me, and all of you blog readers.  Oh, and now I have to kill you.


    Bugger bugger bugger

    5.  This was another mill problem.  I had changed the tools to a 1.6mm drill bit, and reset the Z axis zero.  Or had I?  Maybe I had neglected to hit <enter> after the reset.  Anyway, the chuck crashed at high speed into the job, impaling and snapping off the drill bit, gouging the steam chest, and the drill chuck gouged the milling vice.  The chuck survived but required some remodelling on the belt sander and then a diamond file.   The vice jaw also needed some impact craters to be flattened, then swapped out to another less critical vice.


    I flattened the gouges in the steam chest face, and I will make sure to fill those with something at final assembly.  The embedded drill bit can stay there, after flattening it with emery paper.


There were some other more minor issues, which do not bear repeating and prolonging this missive.  This all happened over 2 days.  Mistakes are made, and I console myself with my father’s advice “he who makes no mistakes makes nothing”.  But, this is the worst run of blues which I can remember.   It IS hot, which is not ideal machining conditions.  So what do I do?

Well, maybe it is just a bad patch, and things will be better next session.  And, I will try to be not SO impatient to get things finished that I don’t double check.  I (and you) will just have to wait and see.

BTW, have a safe and happy new year. 2020.  It is 101 years since we had a double number year (1919), and most of us will not see the next one (2121).  So make the most of it!