How to change cylinder diameter in ST? 1

[mclaser]

It's got to be simple but can anyone tell me how to change a cylinder's dia in ST? 1) make a circle, ESC, 2) extrude to cylinder. Length change is no problem with wheel, but what the heck do you click on/select to mod the diameter? I know you can add a PMI dimension and change it there, but that does seem to fly with a more complex model.

Greg

 

[teebarr]

bc
On the ST release notes under "Adopting ST" it does give details of which functionality still requires the trad. mode.
In reallity it takes some time to get used to the new modelling concepts so I don't expect anyone will start to model complex parts from the beginning. By the time you have become familiar with ST mode you will already know what tools are available.
I think the expectation is that there will be a gradual migration to the new mode as the functionality gets fleshed out. In the meantime it is definetly great for 70% of what I do.
Tony

 

[mclaser]

bc,

I searched through the release notes for the details you spoke of. The closest I could find was

For certain classes of work, you may wish to make use of traditional assemblies, although they can contain synchronous parts. For example, if you use XpresRoute, Frame, or Wiring, you will want to do these in a traditional assembly at this time. Likewise, when cutting assembly features into a part or adding weld material, a traditional assembly is required, though again, you can cut or add weld material to synchronous parts in the context of that assembly.

That's not exactly what I would consider being forthcoming (or detailed, for that matter). Is there another document you're referring to? I've been using ST for 2 weeks and still seem to keep running into surprise limitations. If there is a list somewhere, I think many would be interested in it. Could you pls provide a link.
Thanks, Greg

 

[teebarr]

Greg
What sort of limitations (other than those listed above)have you been experiencing?
Tony

 

[mclaser]

I'm a little reluctant to enumerate issues I've had or seen. I've kind of put ST on the backburner as I can't really be productive with it and have been monitoring the UGS group to revisit the pgm after there's been an update an hopefully cleaned things up. I think I like ST, certainly its promise; however, it's not close to being productive for me at this time. I believe that I've given it more than a fair shot.

Ok, here are a few (from memory). Please don't beat me up point by point. I'm just trying to give others my impression. Yes, I might not be entirely accurate and there might be workarounds, but these are my impressions after using/researching ST for over a week. I would encourage others to add to the list. Maybe a new post would be in order?

- No Divide Part. One could probably workaround with Slice.
- No Inter Part Copy (Part)
- Many operations, esp. forming new solid shapes are not undo-able. There's often no way to get back to the previous condition before your last function. MAJOR time burner.
- Several other functions are not undo-able that need to be. I find my self frequently looking at the undo icon (with a hopeful look) and being pissed that it's greyed out after performing a function that didn't do what I thought it would.
- Dimension anchors (1 side is fixed) aren't sticky.
- Pattern input boxes flash and cycle and severely lags input response. Apparently uses a different user input mechanism than every other function.
- Data input boxes are not consistent. Some don't respond to RMB for Accept. Entered values are not sticky. Round in particular.
- Interface is BIG and needlessly waste screen space. It's poorly designed requiring many more clicks than v20. Autohide is supposed to be the solution but it doesn't automatically re-expand on invocation. See ASM below.
- In Part, HOME tab has most sketching functions but not all. I still need to switch to SKETCHING tab to locate remaining functions (include, move, convert to construction, etc.). Extra clicks.
- In ASM, you can be in a function like Assemble and the active tab doesn't automatically come to the front. This happens even when the pgm is waiting for user to input data in a box within. Again you must CLICK to see it, after you're realized where it's hiding.
- PMI dimension size is proper for only a pre-defined (not by the user) model environment. They don't scale with user zooming. Yes, one can create several new Styles but again, extra work. Click, select size; zoom, click select new size.
- Dimensions are difficult to reposition. It's not clear where to click to move in some cases. Sometimes I've tried 1/2 doz times or more to find the secret spot. Then give up and use the Pathfinder. Pattern label in particular.
- Clicking on items in the model doesn't always highlight the item in the Pathfinder.
- Quickbar doesn't seem save me much time, if any. You still have to move down through the list to find your selection. The functions options are already display (taking up more screen space) on the left side. As all the rest it's not sticky -- doesn't remember your last selection. Why a requestor for something that's simply bi-modal (ie. Add/Subtract)? If you're in Add, a click should switch you to Sub.
- No apparent way to modify (edit) solid constructed by using curved profiles. You can stack, either Add or Cut, another solid to the existing model, however, now you have additional complexity (more items in Pathfinder). The material you just added can not always immediately be detached/deleted. (Probably why Undo is disabled.) The curve nodes show up but there's apparently no way to select them, and therefore modify the shape.
- Misc modeling/display bugs.

Bottom line is it seems to take forever to perform simple tasks. Click, click, click...click. Yes, it's great fun to stack and cut a bunch of cubes and circles, move coplanar faces, etc., but I need sticky dimensions, sticky relationships, editable shapes, Inter Part relationships, Divide Part, UNDO!!, dimensions I can see and manipulate -- these things aren't refined at this time and end up taking much more time than the old methodology. Just my opinion.

Greg

 

[donyoung]

just came across this:

Entered values are not sticky. Round in particular.

for Round values to be sticky you have to switch OFF the dynamic
preview within options -- view otherwise it will always start with
5mm (metric template) This is also is applicable to V20.

Many major functions like DivdePart are planned for the next version :-(

Some functions work different in ST just because there is no history
to back up to when a feature is deleted and thus the Delete or
Undo is impossible (draft is a fine example for being not deletable)

The Sync part IMHO is an unfinished product brought to market to early
"Banana software, picked green, will ripen at the customer's site"

dy

 

[beachcomber]

My colleague has just done a 2-day ST update course.
He's not impressed at all with the usability of ST, and neither were most of the people on the course.
He's an experienced user and, like myself, will generally not follow the training course notes to the letter, and will push the system to see if it breaks.
That's because we all work differently and the software should be capable of doing this.
Since the course he has loaded ST on a laptop and had a play. Some of the things he has shown me has made me cringe.
I know it's V1 but IMO, and taking into account all the hype, this is a really bad release.
Model a cube, put a cutout through it from the front face. Now put another cutout from the top face that intersects with it. Now change the depth of the top cutout. You can't do it because there is no bottom face to move. You have to start filling it in.
Come on Siemens, this is probably the first model anyone trying a new system would try to create and modify, and it doesn't work. This is REALLY poor.
Don't get me wrong, there is some really clever stuff in there. BUT from what I can see it doesn't do anything you can't do already, although it may do things differently, and in fact it does a lot less.
He isn't impressed, I'm not impressed and the boss certainly isn't impressed (he is a user also).

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

 

[teebarr]

bc
You must leave your traditional modelling techniques behind. This works in a TOTALLY different manner. Until you accept that, you will always be frustrated. To learn you must follow the tutorials thoroughly. ASSUMING you know how it should work will fail - it is just too different.
After a couple of weeks (not days) you will start to see the potential of this new way of modelling.

I find it sad that people are keen to slam something they have clearly not bothered to try and learn. Many of the comments show blatant ignorance of the product rather than informed critisism.

Tony

 

[donyoung]

Tony,

even when you leave all behind -- disappointing and far from
being used in a production environment -- unless your parts
are very simple ...

dy

 

[beachcomber]

Tony,
I'm not just slamming it because I don't understand it.
If you introduce a technology like this, with the claims that were made, it has to work out of the box. You just can't make really simple things MORE difficult. You have to make the complex stuff easier.

Don.. How much simpler than a cube with 2 cutouts ?

From a personal point of view, if I'd been given this to try for 30 days I would have done that test first. It's probably the easiest thing to do on any modelling system.
Cube > cutouts > modify.
If I had to start adding in material to blank out other features that I SHOULD be able to modify (otherwise what's the point of any kind of feature list), I would have thrown it out - as I'm sure some potential customers will.

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

 

[mclaser]

(dy, from above) Thanks for the tip, Don. Don't know how your figure some of that stuff out.

I think ST has great promise. See

http://vimeo.com/1725889

for excellent example.

I agree that ST works quite different than a lot of us are probably used to --- that's not necessarily a bad thing. I think the problem lies in the implementation and control. The interface isn't well thought out. All of the click-click-click crap (and wasted screen space) is typical Microsoft and a sign that Siemens probably decided that they could save manpower by detaching themselves somewhat from the UI. Well, the UI is how one drives their application. When it's clunky it doesn't matter what's going on below it, the experience suffers. In an application like this, software designers should NEVER require a user to click three times when 1 or 2 will suffice, or click to un-hide an input requester.

Removing a function like undo is crazy. I have ALWAYS cussed every single program I have ever used that removes or limits this critical function. And you can't tell me (we'll, maybe you can) that they are unable to implement undo just because there is no longer a history tree. The program knows it's last state; it's a simple matter to back up as far as you want to. Big deal, so it takes up more memory. I bet 99% of us would vote for using more memory instead of losing undo capability.

There are other issues but I think there would be less dissatisfaction if this release would have been presented for what it is, not for what it may become.

Greg

 

[donyoung]

Greg,

Undo is not removed in general. Why you sometimes can't Undo maybe
a bug or something else. In any case it should be clearly documented
why and when it is not available.

The other thing is Delete which will not work for certain functions
i.e. applying a draft to a shaft/face of a block. The draft will now
replace one face by another so the original shape is lost In history
mode it is just a boolean subtraction so the original is still there
but in non-history mode this is no longer the case. The draft can be
seen as a milling operation. After removal of the material there is
no way back.
This behaviour is common to some functions we expect to work in
a certain way in history mode but that expectation can't be carried
over to non-history mode.
OTH we are now in a situation where we have nothing to compare with and
therefore do not know if something works as designed or not -- that's
our real dilemma now: to file an IR or not to file that's the question. I fully agree with your final statement.

dy

 

[teebarr]

bc
Kind of proves my point.

"If I had to start adding in material to blank out other features that I SHOULD be able to modify (otherwise what's the point of any kind of feature list), I would have thrown it out "

If you were used to using Spaceclaim of Co-Create you would not find this workflow as odd. The only reason you think it SHOULD be modifiable from the tree is because that is how you are used to doing it.

In ST the tree is only used to select faces not to edit definitions - this takes some time to get usd to. Remember, you are always editing the model, not features via the tree.

The best analogy I can think of is the transiton from horses to cars. In the early days there was little reason to buy a car because a horse could do most things better. In time as the technology improved the tide slowly changed. See V1 for what it is - V1 - just like every other V1 of ANYTHING - limited and in need of refinement. Personally, the more I use it the more impressed I am by what the deleopment team have achieve in terms of useability - it is very close.

I understand it doesn't yet do everything V20 can, but being frustrated because you can't find where to fit your saddle is not so helpful.

For what is worth, at first I hated the UI but now I like it. Give it time.

Tony

 

[donyoung]

... I'm not convinced neither of the GUI nor of the 'synchronous'
in ST, nothing against non-history modelling but not the way
it's now. OK it's V1 but I've to pay for something I can't
use in a production environment. And I fear that all the
backlog of unimplemented ERs will not find it's way into
the TE part of SE which I'll have to use for a long
time :-(

dy

 

[Gemnoc]

All right, I admit I succumbed to the ST hype in the recent months. I was eager to have it my hands. Now, I read this thread and it got me thinking.

Direct editing tools have existed for many years. CoCreate and SpaceClaim have been mentioned.

The question I'm asking is, why this sudden departure from history-based modeling? Why now? Is it because Siemens truly believes that ST is the next "evolutionary step"? Or that SE was in a dead-end, and couldn't continue to compete with such strong (as in sales figures) competitors, so they decided to try something else to differentiate themselves from the competition?

I know, the traditional environment is still available. For now. But if I remember correctly, I read something from Dan Staples saying that eventually, SE would discard it.

I hear that Dassault has something similar to ST in the latest V6 release of Catia. It would be interesting to find out what their users think of it.

 

[beachcomber]

I've tried SpaceClaim and Co-Create and Google Sketch-Up.
They are all history-free modellers and in my opinion the models can be more difficult to modify than in SolidEdge or SolidWorks.
I will try ST 'properly' if I can get time, but it just would not do the job I'm on at the moment. The company I work for does not generally have to swap SE data with anyone else so upgrading and sticking with V21 with the addition of a couple of modules now but no further upgrades (thus saving annual maintenance on 4 seats) is a real possibility.
To me it's like a backward step to Catia V4.
I think Siemens have made a BIG mistake releasing this in it's present state. Dassault and AutoDesk will soon use the restrictions and failures to sell even more of their products. What Gemnoc says about SE being in a dead-end is EXACTLY the comment made in a thread on the SW forum. Personally I think it may be that SE (and SW) are reaching the limit of functionality that they can offer. There really isn't that much you can't do with them and they will soon really start to encroach on NX etc.
A far better plan would have beeen to introduce a V21 that purely had performance enhancements and all known bugs removed, perhaps with one or two new features (I can think of some quite easily. If you read the SW forum there are many complaints about performance issues, and this is one area where, I believe, SE has the beating of SW.
I am a massive fan of SE and have used it since V3

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

 

[teebarr]

Just found this quote from Henry Ford and thought it was funny:

"If I'd asked people what they wanted, they would have asked for a better horse"

Tony

 

[donyoung]

and transposing this into our time it would certainly read:
"... better software, that is stable and where marketing gags
have been dismissed in favour of functions that those out there
in the trenches really need"

dy

 

[beachcomber]

Cribbed this off the SW forum - hope you like it !

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

 

[Incinerman]

Having read these posts I feel better! I have been using Solid Edge extensively to design gasification plant, so its all flanges, pipes, sheet metal and framing. On investigating modelling in the synchronous environment it was as if I had returned to novice status! Some simple models seem nigh on impossible - try creating a hexagonal plate, constrained so that you can change the AF dimension and have it retain its shape. Damned if I can!

 

[KENAT]

OK, I hadn't been looking at this thread but as it's decended into what we think of ST, I've got to say I'm sceptical. We had our rep out doing the dog & pony show couple weeks ago.

If all I wanted to do was make pretty 3D models then great.

If I want to use it to create robust definition of a item, with a dimensioning & tolerance scheme that can be used for inspection etc. then I'm not sure how well it does it.

Maybe I haven't got my head around it yet but I have concerns, this on top of the fact it's almost like a new software, you need to train users, and I don't mean a couple of 1hr lunch and learns, your company 'experts' are no longer experts, any best practices or standards etc you've developed may need replacing...



KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at

http://eng-tips.com/market.cfm?