I have a room of which several parts are hollowed out from one side to the other like pipes with external and internal diameter, the internal diameter being analogous to my recesses When I make an impression with recess, the removal is done correctly on the periphery of the part, But on the hollowed out parts the retraction is not done inwards but always outwards in the same direction as on the periphery. How to remedy this. Second question, is it possible to make fingerprints with draft and also removals with drafts. Thank you in advance for your precious help.
I didn't understand the history of the removals, but I can already give you an answer for the second question concerning the remains.
So yes, it is possible to make drafts in your footprints (see images above).
The "Strip outward" box allows you to change the direction of the draft.
PS: I used SolidWorks 2021 SP5.1, on SW 2020 it works too, I couldn't look on older versions. On SW 2019 and below I couldn't open your file so I'm assuming you're working on SolidWorks 2020.
I understand your disappointment and it is probable that my answer will not suit you: but it is nevertheless a simple explanation.
All Solidworks is based on extruding or removing material from a sketch, which is attached to any plane or face of the model.
In your case you don't use the same reference for "draft nut" and "draft bearing" which is why you can't have the same settings.
Explanation:
The draft is always done from the sketch and more precisely from the sketch plan. Which is logical for those who make single- or two-part molds. But you start with a sketch inside the gear for the bearing drafts and an outer surface for the nut draft. The outer draft will therefore be done upside down (undercut) while the others being inside, the removal will be done outwards as standard.
That's why you are forced to reverse the direction only for the cap thanks to this function provided precisely or someone would not use the same reference frame. Note that instead of reversing the direction you could shift the material removal plan but this is more cumbersome and risky than reversing the direction!
I don't make molds anymore but I used to start from a single frame of reference starting from the separation plane for single, two-part molds. For multi-part molds it is a little more complicated but the principle is the same.
Kind regards
[HS On]
I meddle in what doesn't concern me so please don't set me on fire ;-)
There is something wrong because if because of the remains it is a priori for demolding, but only some parts have a draft.
I conclude that it is not for demolding since all the other surfaces have not stripped but to facilitate bearing or nut fits
This can be understood for the bolt that can be assumed to be very tight, it is less understandable for the bearing stop, especially if it is in the center of the gear. I don't see how you can insert it. I have other remarks on the size of the draft chamfers but I don't want to abuse your patience ;-) so I abstain ;-)
You probably use the Fingerprint part function ... to rough a mold of the part, taking into account shrinkage on cooling. The dimensional variation is assumed to be linear, of the form L' = L (1 + k * dT). For SolidWorks, the term k * dT represents the percentage increase in dimension. For a thermally isotropic material, it is applied uniformly in all directions in the vicinity of a point. SolidWorks offers the option to distinguish values according to the X, Y, and Z directions of the coordinate system.
In the case of foundry, shrinkage can be compensated by an enlarged footprint of X%. In this enlargement phase, the recesses of the room move away from the center, which seems to bother you, but is very normal... To be convinced, you just have to imagine a ring that expands under the effect of a rise in temperature, the outer diameter increases, the inner diameter as well, but slightly less...
.
As for the drafts of the model, they can be set up when defining the extrusion functions of the part, by choosing the direction of the draft appropriately. It is probably excessive on the attached model (SW 2020), in order to better distinguish dimensional variations.
