Change cylinder diameter in ST

[beachcomber]

Is here a way to change the diameter without adding a dimension?
All I seem to get is move or copy.

bc.
2.4GHz Core2 Quad, 4GB RAM,
Quadro FX4600.

 

[donyoung]

BC,

ST-MOde? I can't see how it should go with Move/Copy, but adding
a dimension will work as well as Offset with Attach (Add)

dy

 

[beachcomber]

Hi Don,
I'm trying to get my head round this ST stuff.
I've done the tutorials but they are very limited.
So I've created a cylinder 90mm diameter x 50 long.
I now want to make it 100 dia x 75 long.
A simple task really.
In SpaceClaim I just click the cylinder wall and up pops a box telling me it's 90, so I change the value to 100 - don't need any dimensions. I then click the end face, the ruler icon and then the other end face and it tells me the distance is 50, so I change it to 75. Job done. No dims to add, everything is done on the fly.
How do I do it in ST ? Are there equivalent/similar actions?
I know in this case I can just move the face 25mm to extend it - but I want to know that I'm making it 75 long, not just 25 longer.

bc.
2.4GHz Core2 Quad, 4GB RAM,
Quadro FX4600.

 

[donyoung]

BC,

ST-Mode is not so easy -- because it's up to 100 times
faster ... So you have to dimesion the diameter as well as
the length to be able to set the value to a well defined value.
Moving or rotating a piece with the SteeringWheel is always
relative so you won't get the current dimension/angle
And remember: dimensioning is now limited to
- endpoint to endpoint
- center to center (arc, circle)
or a combination thereof. Other dimensions might be placed
but you can't change them.

dy

 

[teebarr]

bc
You are right, the only way to drive geometry is to add dimensions. ST has taken a different route than other non-history systems with a view to giving you some of the power of a history system. When you add a dim. it appears in the variable table and can now be subject to design rules, formulas or driven from a spreadsheet.
So while there is an extra step, there is a good reason for it. There is (like in trad.) the option to automatically apply dims during sketch creation.
While ST has many things in common with SC, it is designed from a different perpsective - the ability to tie down geometry and capture design intent.

 

[donyoung]

> When you add a dim. it appears in the variable table and can now
be subject to design rules, formulas or driven from a spreadsheet.

automation in this way has gone. The variable table does not
give you the information in which direction a change in value
will go. When a change in value can't be solved there is no
indication of that error. Any dimension used in formulas
must be a fxed dimension
The way SpaceClaim handles dimension changes is a lot better
I think. I personally can't see no benefit in creating a dimension
just to have the opportunity to change a model when the same can
be done without.

> There is (like in trad.) the option to automatically apply dims
during sketch creation.

not all dimension done in sketch-mode will be propagated to
the final model, some will only have informal character because
you can't change them

dy

 

[teebarr]

dy
I am at a bit of a loss as to know what you mean. It is possible that my models are quite simple so have not encountered the problems you mentioned. Maybe you could give a specific example so I can see what you mean.

>automation in this way has gone.

I don't understand - I am using it. It is possible there are limitations I haven't come across yet. If you had an example it would be helpful.

>The variable table does not give you the information in which direction a change in value will go.

How is this different to trad. From what I can see it responds in exactly the same way- it will move in an unconstrained direction.

>Any dimension used in formulas must be a fxed dimension

Again, this is simply mimicking driving dimensions in trad. As you know, in trad. you can not use a driven dimension in formulas - only driving.

>The way SpaceClaim handles dimension changes is a lot better I think.

By implication you are saying you prefer the limitations of a standard non-history modeller like SC and co-create - ie the inability to drive geometry externally. There is nothing wrong with that, this suits many people. But it then begs the question why you are using a history based programme in the first place when this technology has been around for a long time.

 

[donyoung]

> ie the inability to drive geometry externally

BC has shown how to drive it externally, he has not stated
that there are no variables

> How is this different to trad. From what I can see it responds in
exactly the same way- it will move in an unconstrained direction.

no, the user can change the direction. SE will use the last
setting of the action arrow. Have a block and dimension one
side and let the arrow point to the right. Updating the var
will work as intendend (maybe). Now change the arrow to point
to the left and change the var. Now the chnage will go into
the opposite direction which might not be the wanted one.
To make an ST-Model behave completely deterministic like
it was the case in trad. mode then ST has no benefit. Also
by using variables in ST-mode the LiveRules have no effect.

Change the attached model from 2.5" to 2" by using variables
only (or anything else that is dimensioned)
The other sample can be changed in diameter, thickness, wave
peak by just using variables.
Both are fully parametric in trad mode and easy to create
and can be changed by just using variables and they behave
completey deterministic

Simple sample: take a plate of 0.5" thickness, round one edge
with 0.5. Now try to change or get rid of that rounding.

I've put ST aside and continue with V20 -- I have to make
money and V100 (trad. mode) is no option ...

dy

 

[beachcomber]

Thanks all.
It's clear that ST may be useful for simple tasks and editing imported models, but even after a very brief trial I find it too unpredictable.
Having to add dimensions and fix them etc to control a model correctly seems to be no quicker than traditional mode.
I also found that when I added a dimension to the end of the cylinder - on the circular edge because you can't dimension the cylinder itself - the dimension becomes detached if you then round the edge. To me this is rediculous.

bc.
2.4GHz Core2 Quad, 4GB RAM,
Quadro FX4600.

 

[donyoung]

BC,

> I also found that when I added a dimension to the end of the cylinder -
on the circular edge because you can't dimension the cylinder itself -
the dimension becomes detached if you then round the edge. To me this is
rediculous

AFAIK there will be a change in ST2. But remember: in ST-mode a feature
actually modifies another feature whereas in history mode it's not
the case. There a feature only overlays another one. So in your case
the face/edge the dimension is attached to is no longer there because
of the rounding. And a face/edge that has dropped off the BREP is
gone. When you delete the rounding the dimension will become attached
again. But you can't delete a feature that has fully 'consumed' a face
on the feature it is applied to (a draft on a face is an example on
this behaviour or the rounding of the plate in my previous post).

dy

 

[teebarr]

>BC has shown how to drive it externally, he has not stated
that there are no variables

I was suggesting the reason ST works differently to SC in this respect is that in order to populate the variable table you need external dims.


Don, thanks for taking time to do the examples. I agree, these are issues but I think these are temporary and as you mention, ST2 sounds like it has gone some way to resolve them. We will have to see.

My point is, we have now been given 2 options for modelling. Each has its strengths. Sync may never be able to do all Trad can, but in its own right it is a good tool and I have every confidence that Dan and his team will shortly make it into a truely great tool.

In the meantime, getting to understand the shortcomings helps you to avoid certain issues. The round is a good example. You also learn to model differently, in the same way you would with SC e.g. instead of deleting a feature you may chop it off.

So far, I have found it great for imported models and working up design concepts. If that is all it can do then that is still better than what the competition are offereing. My belief however is that it will develop onto a fully formed modelling tool.

I agree though that it is not fully formed enough to do most of your modelling tasks. Also, you need the luxury of time to learn it. Siemens have not done a great job in providing tutorials that give a good understanding of how the thing works. More detail would help.

In this new project I plan to model some "real" parts in Sync - I'll let you know how it goes.

Tony

 

[donyoung]

> I agree, these are issues but I think these are temporary and ...

I don't think so. That's the way it works now. Remember: a feature
is now only a collection of faces and nothing more. A coutout put through
a square tube results in two independend cutouts. Each can be changed,
moved or even deleted without affecting the other.


Here is another fine (also simple) example of problems that one
has to be aware of:

LR set (shortcuts): C, P, T, L, S(all), A, O

- B is variable, A is fixed
changing B in either direction will not work. But this is
not prevented by dim A wich is fixed, here the LiveRule
'Orthogonal to base ..' does prevent the change. Deactivate
and it will work

- B is fixed, A is fixed
same result as above

- B is fixed, A is fixed
changing the *angle* in either direction will work. When the
direction of change is toward the 'B' face a rhombic shape
will be the result. Reversing the direction the 'B' face
alone will move

- B is fixed, A is fixed. Additional LR set: Perpendicular (D)
changing the *angle* with the direction of change set towards
the 'B' face - the change is not possible. Reversing the direction
towards the (unlabeled) front face - the 'B' face will move on
it's own.

Apply this to a complicated part combined with a very tight
schedule to complete a change and the data will go directly to
the CAM ...

dy

 

[donyoung]

/Edit

forgotten to mention: there is a solution to change just the
width of the 'rocker arm' from 2.5" to 2" but there is none
for the wave washer other than delete and recreate it.

 

[teebarr]

Don
Another good example and explaination.

My experience of driving geometry from the variable table in Sync is limited to some simple changes to imported models of bearings and housings etc. It was very useful and effective.
But I see what you mean, the ability to drive geometry in the way we are used to in Trad. is limited.

>these are issues but I think these are temporary

My guess is that in the future, procedural features will be refined to get around mant of these "fundamental" issues.
The round issue is a good example. Fully rounding an edge so that it cannot be edited is clearly not acceptable - that is why I am certain they will find a way to refine this to make it work.

On the other hand, your examples show the flexibility that this type of modelling allows. What I am interested to find out is if I can understand it sufficiently to make it work for me. As I have mentioned, I have already been using it this week for design concept work. This has worked well and I have enjoyed the freedom to develop ideas quickly. As the design progresses, I plan to model some parts for real in Sync to see if the flexibility I enjoy at concept stage turns into frustration at defined part stage.

Understanding the weaknesses of this technology is essential to avoiding frustration (as with any modelling software) but understanding it's strengths is equally important if we are to exploit it's inclusion in SE. At this stage I don't see it as a relacement to Trad. but a compliment. In that light it has the possibility of helping the design process.

Tony

 

[burhop]

Hi Beachcomber,

I don't have ST on my machine anymore but I do have ST2 beta. I took a quick look with both holes and cutouts and it seems to work they way you want it to in this release:

http://screencast.com/t/Gvp6dsNzoyk

 

[lapuser]

Whats the latest rumor on a release date for this?

 

[Gemnoc]

If V20 and ST are any reference, should be end of August/early September.

 

[donyoung]

official statement was: mid summer ...

 

[karmoh]

Or mid-winter...depending where you are

 

[jdg268]

Just to tag along to the original question, ST2 gives you the capability to edit a diameter/radius via a specific value, whether you placed it there or not. You can't change a diameter/radius by dragging the steering wheel though.

Last I heard was in a few week, by the end of August...



John Graham
Applications Specialist
Designfusion