How to scale size of items BUT leave the center points of all scaled items in the exact same place

I’m working on a custom image press that has male and female halves. A sheet of material will be placed between them and the two halves will be pressed together, embossing the image onto the material.

I can easily make the male and female halves that fit together exactly. Just make a copy of the male design, insert it into a blank soon-to-be-female block, and then use a boolean difference to remove the male.

BUT that doesn’t account for the thickness of the material that I want to insert between them. So I need to male piece to be scaled down a smidge. The scale and scale2d commands don’t seem to work because they MOVE the locations of the image. I want the objects on the male block to be smaller IN THE EXACT SAME PLACE they currently occupy. More accurately, their center would be in the same place and their height and edges would be contracted a bit from where they originally were.

Is there a way to do that?


Hi David,

It sounds to me that, rather than scaling the male piece, you would get a better result if you offset its mating surface(s) by the thickness of the material (or perhaps slightly less).

That would allow for a consistent thickness across the image. Scaling would result in variable thickness depending on the orientation of the surface normal at each point.



That’s a good idea. Here’s the problem I see with it. It leaves room for the material at the “top” of the male piece, i.e., where it’s deepest into the female piece. But the sides of the male and female piece would still want to touch where the male piece is taller than the material is thick. There would be no room for the material between them and it could cause a sheering cut thru the material. (That’s a concern with the material I’ll be using, think aluminum foil only out of fine silver and a bit thinner.)

Hello - I am not sure if I am picturing things correctly but is seems to me you can make one half of your press, OffsetSrf the entire set of faces that will have material between them by material thickness, and make the other half from that?


Here are a couple of pictures to illustrate the difference between scaling and offsetting.


Note the inconsistent gap.


The gap is much more consistent.

But note that OffsetSrf works best when the offset distance is smaller than the local curvature of any point on the surface. It isn’t happy with bigger offsets that result in self-intersections.

dear @david_wendelken
you may want to accept Pascals answer as solution ? or comment what you still miss.

I was a bit mislead by your topic:
" How to scale size of items BUT leave the center points of all scaled items in the exact same place"
the answer to this question would be
using the option “transform individually”.

but as others already mentioned your description shows you re not searching a scale operation. Maybe the topic is more something like:
how to handle / add sheet thickness for positive / negative mold ?

you may want to adjust the title ?

kind regards and happy modelling. - tom

Tried getting the offset surface approach last night. I think it might work for me but I was having some trouble with it. I’ll try again today and go thru some tutorials on how to use it correctly in order to be sure.

And thanks everyone! I’ve only been using this tool for a couple of months and there’s so much to learn!

I’ve tried the OffsetSrf command quite a few times today and it’s almost working for me but I think I’m doing something wrong.

Here’s what I’m starting with:

I start with an image, turn it into curves that have matching, offset curves, then extrude those curves into solids. I then place the solid I just created onto a solid box. You can see the progression of the different steps in the image.

I found that I needed to have a pretty small offset or the program got confused due to the intricate designs, much like the fillet series of commands can do.

The closest I’ve gotten it to work was when it created an offset of the entire solid I created in all directions around the solid box with the solid “wire” pattern. Couldn’t get the boolean union/difference trick I’ve used to make exact male/female matches to work (plus in this case get rid of the unwanted “wrong side offset” .

Any idea where I’m going wrong?

Hi David.

I don’t know how you get those curves, but looking at your screenshot ( green ), I am wondering if
complexity is not the root of the expressed difficulties ?

You may have few invalid curves or contours that are too small for the intended purpose…

Did you tried with portions of the set of curve instead of using all of them ?
If that works, it may help to understand what’s going on there.

Rodolfo Santos

I’ll give it a try. Thanks.

Hi David - we’d really need to look at the geometry in the file- it looks like there is plenty of scope for ‘issues’ - please post that or send to along with a link back to this topic in your comments.


dear David
what is the final application / target use of your data ?

if it is milling for example - there might be easy solutions where you do not need the corresponding data … but can “fake” it in the CAM-Package or on the machine side.
…for example you can generate machine-pathes with a radius R4 Ball-Mill but put a R5 into the real machine - to remove 1mm of material… (z-level is unimportant as it is the press-direction)


kind regars -tom

I’m looking to make an embossing press with custom designs for very thin silver foil - think aluminum foil or thinner. I’ll be enameling over the foil, which will add a very nice surface texture and look below the transparent enamel.

Ok, everyone. Thanks for the input.

I’ve done some experimenting this week. It looks like the offsetSrf command will work well for simpler designs. More complex curves with tight corners seem to confuse it, much like the fillet command gets confused.

(pictures uploaded in opposite order described. oops.)

I take that back! I have a very simple “O” shaped design. Here’s a rendering of the original item followed by 3 offsetsrf results, with rounded corners and an offset distance of 0.01, 0.02 and 0.10 mm.

As the offset distance gets larger the gaps in the created object get bigger and bigger. Yikes!

Took me a moment to work out that your pics are back to front…

Could you post the geometry for the original shape?

Had to make a copy of the file, I guess the original was too big to upload.

test offsetsrf.3dm (82.6 KB)

Hi @david_wendelken,

I see what you mean. However, I was able to get a clean offset by fairing your original lines once and recreating the ring from them. Unfortunately, I’m not smart enough to tell you why the offset doesn’t like the original rings - the faired curves are not very different from them. Hopefully someone like @davidcockey or @pascal could shed some light on that.

test offsetsrf-faired.3dm (840.0 KB)

Wow! That’s way, way better! I’ll go learn about fairing and see if that doesn’t solve the problems with the more complex curves, too!

So much to learn!

And you are all the best!