Using image-to-gcode - tool diameter problem

I'm designing a bridge for a musical instrument.
I've generated a greyscale image in CorelDraw, consisting of a blend of two elipse set in an outer rectangle, exported it as a gif file, then opened it in Axis.
All OK, except the outer edges of the tool path, and the overall dimensions of the piece are smaller than the design by a size equal to the tool diameter.
Please see attached files.

The finished piece will consist of a set of nine of these, in a line like a mountain chain, and I had assumed that I would run a repeat command of some sort, rather than start with art work that had all nine on it.
Or is this what I will need to create ?
Any thoughts ?
John

Making a 9 unit version was easier than I thought, and the 3500 lines of code take about 40 minutes to run, so as I shall be cutting hard wax for a model, that's acceptable.
The dimensional change is still only one diameter around the whole piece, so I assume I just need to get my head around how it works.


John

you have not set an outside offset to the toolpath , which is the radius of the tool used

Ah, good point.
However, when I opened the .gif file, one of the various settings in the window that opens was "Tool diameter" which I set to 4mm, so I assumed that was it.
I've just gone through the process again, and my original settings are still there, with the diameter correctly set.

??

Regards
John
EDIT I've just looked at the gcode in the file that the software produces, and there's no indication that a particular size for the tool has been set.
Should there be, and what would the code read ?

the overall dimensions of the piece are smaller than the design by a size equal to the tool diameter.

You might just scale your model in Corel to be a larger size (parallel line around the object (contour?) by 1/2 the tool diameter as it appears the outside path c/l is following the outside of the design.

Rick G

Very true, Rick, and indeed that is what I've done in adding a small rectangle outside my original surface model.
However, that doesn't teach me anything about why the software behaves in this way.
A bit like a doctor treating the symptoms rather than the cause !
Regards
John

Reinforced by coffee and toast, I think I now have an explanation of my problem.

First, the Corel program let me generate a greyscale map of what amounts to a Z axis only view of my model. This is a series of filled ellipses, going from white in the centre to black outside.
If the "blend" that Corel performs is made with no "acceleration", the result is a linear step size from each ellipse to the next, so the profile seen from the side would be a straight slope.
Having chosen a high acceleration to the blend, the resulting profile is a truncated bell curve, which is what I'm aiming for.

When the saved .gif file is opened by Axis, it requires a set of parameters to be input, and it then generates a display of the resulting tool path.

Two things have now occurred.

The first is that the tool path doesn't allow the tool profile( generated from tool diameter and tip shape input as above) to cut into the shape being generated. I've no idea how it does this at present, but just accept that it does !
This means, in my case, that if the surface being generated has a curve of smaller radius than the tool, which I suspect may be occurring near the bottom of the bell curve, the path will avoid this conflict.

The second is that the path shown on the Axis window appears to show the shape of the object.
It doesn't.
The image of the tool path looks very much like the finished object, but it will be "fatter" around the X-Y plane by an amount equal to the tool diameter.

I think these two effects merge together, and have given me the present problem.

As my judgement of the shape is purely an aesthetic one at this stage, because there is no reference grid available as a background, I have been taken in by the similarity of the tool path to the shape I'm trying to produce.

It seems to me now that the only course is to cast some wax, and get cutting.

This in turn now makes me wonder about the choice of "alternating" path when cutting wax.
Will it be liable to melt when the tool is cutting in the "wrong" direction ?

Ho hum.

i presume your trying to produce in 3d , you would probably be better with a 3d drawing and then processing the stl file into gcode ...

which wax are you using ? , i tend to use ball end slot mills with wax , or in general slot mills , i use trycut wax

Dave

Hi Dave.
I'm using Corel as I have been using it for about 15 years, and feel a bit long in the tooth to start over !
As for wax, I've got about 20lbs paraffin wax with additives from days gone by when I used to make my own oval picture frame moulds.
I built a hand driven Archimedean oval lathe chuck about 2ft by 3ft, and used fixed profile tools to scrape the wax into the shape I needed !
So once again, I'd prefer to stick with what I'm familiar with.
It has occurred to me to do a first test with MDF anyway, but another problem has surfaced(sorry).
My ball nose mill hasn't got enough cutting length, so I recoded with a roughing pass.
Ho hum again.
John

Edit Just googled trycut wax - looks very nice - might well get some. Thanks for the idea

ha yes i know exactly what you mean , i dont have any problem with the wax melting as this stuff is quite stable ,
and used for lost wax casting etc , redundant prototypes get remelted into new blocks again using home made aluminium moulds
i use a deep fat fryer to melt the wax .