USER FORUM

*(you are viewing a thread; or go back to
list of threads)*

**flat pattern**

*(by wolfv)*

I made a .slvs file of a folded cardboard mockup.

It has multiple extrusions, where each sketch is constrained to a previous extrusion.

Is there a way to import extrusions from the file, one extrusion at a time?

I want to assemble the extrusions into an unfolded flat pattern.

Thank you.

It has multiple extrusions, where each sketch is constrained to a previous extrusion.

Is there a way to import extrusions from the file, one extrusion at a time?

I want to assemble the extrusions into an unfolded flat pattern.

Thank you.

**(no subject)**

*(by Jonathan Westhues)*

No. You could instead create a part per face, and import them both to a folded assembly and an unfolded assembly.

**constraining the sketches**

*(by wolfv)*

If I create a part per face, is there a way to constrain each sketch to a previous extrusion?

I am assuming that "a part per face" is the same as "a file per face".

Seems like a model can be either flat pattern or parametric, but not both.

But maybe I am missing something.

I am assuming that "a part per face" is the same as "a file per face".

Seems like a model can be either flat pattern or parametric, but not both.

But maybe I am missing something.

**(no subject)**

*(by Jonathan Westhues)*

Correct, file per face. You can import parts into other parts, suppress the imported solid model, and use the imported part just to constrain against.

**(no subject)**

*(by wolfv)*

That worked! Got myself a nice flat pattern.

What is the best way to constrain a sketch to a linked part?

When a new empty file links to a part, the linked part appears off to the side.

I created and constrained the new sketch to the linked part where it was.

Would it be better to constrain the linked part to the origin?

What is the best way to constrain a sketch to a linked part?

When a new empty file links to a part, the linked part appears off to the side.

I created and constrained the new sketch to the linked part where it was.

Would it be better to constrain the linked part to the origin?

**(no subject)**

*(by Jonathan Westhues)*

You should always fully constrain the imported part. If you don't, and you accidentally drag to translate or rotate it, then all of the geometry constrained with respect to that part will need to get re-solved. That's at least confusing, and might in some cases cause the solver to fail, or to find an unintended solution.

Since the only point initially available is the origin, those constraints must involve the origin, like by choosing a point on the imported part and placing it at the origin. You should also constrain the rotation of the imported part, like with "same orientation" on a coordinate axis.

Since the only point initially available is the origin, those constraints must involve the origin, like by choosing a point on the imported part and placing it at the origin. You should also constrain the rotation of the imported part, like with "same orientation" on a coordinate axis.

**How to constraint this part?**

*(by wolfv)*

Sometimes I can not figure out how to fully constrain parts without over constraining.

The attached files demo_constraint.slvs has a linked part over constrained.

Is there a way to constraint the linked part without over constraining it?

Apparently I need more than spacial intuition to solve constraints.

How do other people learn it?

Thank you.

The attached files demo_constraint.slvs has a linked part over constrained.

Is there a way to constraint the linked part without over constraining it?

Apparently I need more than spacial intuition to solve constraints.

How do other people learn it?

Thank you.

**Some first hints ...**

*(by Roland Frank)*

There is not really an optimal workflow.

Just some thougths for guidance.

>Constraining sketches

Start with obvious gemetrical constraints.

Like horizontal, vertical, point on object.

Continue with connecting geometrical contraints.

Like tangential.

Continue with think about geometric relations.

Like equality, symmetrical.

Use construction geometry to help.

For example a horizontal line for co-linear objects.

Set up dimensional constraints.

Prefer horizontal and vertical length over absolute length.

(lesser CPU usage)

>Constraining parts in an assembly

With the few projects I did already i made best experiences

using

- an "on point" constraint to make two points coincident

(one point on each part)

- using "same orientation" on two vectors

(one vector of each part)

or in other words:

- Where to insert part / join parts ?

- Which direction/orientation for second part ?

Hope that helps for first directions ...

Roland

Just some thougths for guidance.

>Constraining sketches

Start with obvious gemetrical constraints.

Like horizontal, vertical, point on object.

Continue with connecting geometrical contraints.

Like tangential.

Continue with think about geometric relations.

Like equality, symmetrical.

Use construction geometry to help.

For example a horizontal line for co-linear objects.

Set up dimensional constraints.

Prefer horizontal and vertical length over absolute length.

(lesser CPU usage)

>Constraining parts in an assembly

With the few projects I did already i made best experiences

using

- an "on point" constraint to make two points coincident

(one point on each part)

- using "same orientation" on two vectors

(one vector of each part)

or in other words:

- Where to insert part / join parts ?

- Which direction/orientation for second part ?

Hope that helps for first directions ...

Roland

**(no subject)**

*(by wolfv)*

Thanks Roland. I forgot about using construction geometry.

So I added construction geometry to the example in demo_constraint.7z.

But it still doesn't constrain correctly.

I submitted an issue on https://github.com/solvespace/solvespace/issues/21

So I added construction geometry to the example in demo_constraint.7z.

But it still doesn't constrain correctly.

I submitted an issue on https://github.com/solvespace/solvespace/issues/21

**(no subject)**

*(by whitequark)*

@wolfv, note that the right way to do the constraining you want is to constrain normals to be in the same orientation.

**Post a reply to this comment:**