MadModder

Some of mesh editing tools in blender have weird ways of working.

So far, most annoying is mirroring an object. The tool itself works as one might expect, but the mirrored object always has a negative scale on mirror axis:

  [ You are not allowed to view attachments ]  

It can be fixed by using 'apply scale' command to reset X, Y and Z scales to positive 1.000.

If the object's scaling is not uniform, it affects modifiers like beveling (which isn't that great in Blender, though). Anyway, I wonder where negative or non-uniform scale can be used without messing things out?

So, once the scale is reset, perhaps because of negative scale value, mirrored object's normals or faces are inverted, which also affects modifiers (such as Boolean operations won't work):

  [ You are not allowed to view attachments ]  

Next thing to do is to select all faces in edit mode, and apply ‘flip’ or ‘recalculate outside’ command.

Same steps have to be repeated over and over every time. It makes otherwise simple operation rather tedious, so I started to look if there are ways to automate previous steps.

One way is to use output of Blender's info window, which shows history of previous actions:

  [ You are not allowed to view attachments ]  

Commands can then be copied to Blender's text editor and executed there as a script. Seems to work quite well, with just one click. But to make a button for it to 3D window instead, I asked ChatGPT - that glorified oracle - how that can be done.

It added some python code to make it happen. Commands included are:

- Duplicate object
- Mirror (in this case in x axis)
- Apply scale
- Enter edit mode
- Selection mode face
- Select all (faces)
- Flip normals
- Enter object mode

Although I'm not much of a programmer, I was able to figure out, what kind of code snippet to add to make a script to an installable add-on:

  [ You are not allowed to view attachments ]  

There are also some other command sequences, that I guess can be handled or 'buttonized' with similar python scripts as well. The script that ChatGPT provided, has comments on what different parts of code does, so it's fairly clear where certain list of commands should be placed.

I've added ferrosilicon routinely to iron melts for the usual reasons. I recently got ahold of some elemental (99.0+ pure) silicon metal, and wondered if that could be used in an iron melt to similar effect.

Melting point is 1414 C , which seems do-able (unlike in a normal aluminum melt). I do realize that the S.G. is much lower than iron, so it will tend to float, and oxidize, and that the silicon effect is transitory, so needs to happen at the very end of the melt.

Other than that, is there a reason why it isn't generally talked about? I mean I just did a search online and after 3 pages of pointless results repeating that (ferro)silicon is used in iron melting, there was nothing to say yay or nay about pure silicon metal addition.

Anybody know -- or even better, have actual experience?

I've been working on the possibility of using a moveable regenerator instead of a displacer, along the lines of experiments done by David Urwick and Mick Collins in the late 70's. I made a metal wool mesh caged moving regenerator with the same overall proportions as No. 83's hollow aluminum displacer. Photo:

  [ You are not allowed to view attachments ]  

 I only got to try it out on the Prony brake for about 15 minutes before the engine seized. The power cylinder was not the source of the seizure -- it was the displacer pushrod in the rod bushing that froze. I'm not entirely sure of the ultimate cause, but I found a foreign powdery substance, (possibly burnt off of the metal wool) deposited on the rod, and it had high enough friction to cant-lock it in the bushing. (Further proof that my bushing is too short, btw.)

While that was easily remedied, I also found tiny bits of wool in the power piston cylinder, and this was a more serious defect, so I called a halt to further running with this regenerator, as built. The photo above was taken after the run, and you can see some of the particles migrating out through the mesh.

While it was running, I did feel like the performance was easier running, but less powerful than the displacer versions. I can't say for sure, since I now wonder what the rod bushing friction was like during the trial.

Top speed was only 966 RPM and calculated power at the only speed tested was 0.134 watts at 935 rpm. That is less than half what the displacers were putting out at about the same speed, but again this isn't definitive since the rod friction had unknown effect at the time. Also, the displacer runs were measured at about a half hour in, when the engine was fully warmed, and showing best results. The regenerator run only got to 15 minutes before seizure.

Still, it's interesting to see the engine run without a displacer, using a moveable regenerator. I think the main difficulty with the latter is actually fabricating one. The requirement is that air be able to pass through it. End caps, of course tend to defeat that, unless perforated, and yet must strictly centralize a pushrod. Perforations aren't ideal, a web like structure would be better. The sides, if mesh, make chucking in a lathe impossible, so you would have to plan all structural machining before assembly.

Assembly also is problematic for a high temperature environment. Silver soldering if steel or stainless steel is used for mesh and end caps is difficult or impossible with finely divided materials, and most glues are tolerant of only lower temps. Press fits aren't possible with mesh.

These requirements would be easier to meet in a larger bore than the ~ 1" diameter of No. 83, so it was something of a victory figure out how to make this one work at all. Unfortunately the wool shedding problem with all the other difficulties leaves me less than enthused about going further in this direction on No. 83. But I'd certainly like to hear from others if they have.

One thing is for certain, they will run on a regenerator alone.

The final version on a wooden base and everything blocked up and clamped in place. I also shortened the stylus considerably:

  [ You are not allowed to view attachments ]  

Well, not strictly machining, but machining adjacent.  I saw this on one of Joe Pie's videos and decided to make one for myself.  I went a little overboard, cutting recesses and then machining screw heads to fit in said recesses.  Main body is made of PVC scraps I get from work.  I used my own DTI and scaled it up 10:1 for the dimensions.

  [ You are not allowed to view attachments ]