Welcome to the Onshape forum! Ask questions and join in the discussions about everything Onshape.
First time visiting? Here are some places to start:- Looking for a certain topic? Check out the categories filter or use Search (upper right).
- Need support? Ask a question to our Community Support category.
- Please submit support tickets for bugs but you can request improvements in the Product Feedback category.
- Be respectful, on topic and if you see a problem, Flag it.
If you would like to contact our Community Manager personally, feel free to send a private message or an email.
how to dimension hole through a stepped shaft?
len_friedland
Member Posts: 13 ✭
Hello,
For my first project in Onshape, I modeled a one inch diameter shaft stepped down to 11/16" diameter. In the area of the step, I passed a 3/8" hole radially through the part.
When I try to pick the hole to dimension, I simply cannot pick it. I can dimension neither the hole diameter nor the hole location.
I even tried adding a point in the center of the hole (by locating the point the correct distance from the end of the shaft) in a separate sketch, but the point doesn't come up in the dimension workspace.
I don't know how to attach the drawing to this post. If someone wants to see it, please provide instructions on how to do so. Thanks.
Could someone please let me know if it is possible to dimension this print in Onshape and, if so, how?
Thank you for any replies.
For my first project in Onshape, I modeled a one inch diameter shaft stepped down to 11/16" diameter. In the area of the step, I passed a 3/8" hole radially through the part.
When I try to pick the hole to dimension, I simply cannot pick it. I can dimension neither the hole diameter nor the hole location.
I even tried adding a point in the center of the hole (by locating the point the correct distance from the end of the shaft) in a separate sketch, but the point doesn't come up in the dimension workspace.
I don't know how to attach the drawing to this post. If someone wants to see it, please provide instructions on how to do so. Thanks.
Could someone please let me know if it is possible to dimension this print in Onshape and, if so, how?
Thank you for any replies.
0
Answers
There are plenty of people here on the forum who are more than willing to help.
Please make the document public and then post a link. If you are unwilling to share your document...screenshots will help.
You can make your document public by clicking the SHARE button at the top right in the document--> then public tab.
https://forum.onshape.com/discussion/2136/start-a-new-sketch-on-a-curved-plane-shaft-with-animated-gif
https://forum.onshape.com/discussion/2139/help-with-a-setscrew-hole-in-collar
Twitter: @onshapetricks & @babart1977
It's not clear from your post whether you have successfully modelled your 3D part (in a Part Studio) and are now seeking to produce a 2D drawing, or are unable to position the hole in the 3D part. Please indicate which of these two workspaces you are having the problem within.
Twitter: @onshapetricks & @babart1977
https://cad.onshape.com/documents/b85499452b334240b03828e6/w/9c8be78133584ae098ab0394
ON EDIT: This webpage would not refresh for me, so I was not able to see the OP's answer to my question about this.
Here's a thread from a few months ago which elicits that info from Onshape.
https://forum.onshape.com/discussion/1681/drafting-dimensioning-holes-diameter-or-radius-in-parts-with-curved-or-spline-driven-surfaces
Hopefully the fix is not far away, but 2D drawings appear to be very challenging to develop on top of a 3D modeller (judging by how long it has taken others to get this facility sorted out and bedded down), and being browser based presumably compounds that difficulty for Onshape and their very capable German 2D CAD module partner, Seibert.
It tends to be much harder to come up with a workaround in 2D modules like this, than in 3D, at a similar stage of development. The only ones I can think of are so amateurish I would be embarrassed to recommend them on a public forum.
Hopefully someone more clued up will save the day.
https://cad.onshape.com/documents/6fbade97d6c14b0198b16455/w/f87d731e3d5540dfb7f2ea9e/e/740289717cb54a70b64b25b2
I RMC on view and showed hidden lines (I made the model translucent but not sure that mattered. Then dimensioned hole picking point to point to dimension. Hopefully that was what you were after.
I was intrigued by your answer, given it had not previously been (in my experience) possible to dimension to hidden lines. I checked out your model (thanks!) and found it is still not possible: your method relies on picking the points where the hole breaks through the respective cylindrical faces, and works whether or not hidden lines are displayed
vs
However your method DOES have the advantage of indicating to the person reading the drawing what is being dimensioned.
(Unfortunately, strictly speaking, most drawing standards forbid dimensioning to hidden detail, which I personally find unduly anal)
The OP still has a problem, in either case, because he still cannot dimension to the centre of the hole to indicate where to machine it. The new ability to provide a centreline between model lines does not work for hidden detail lines.
I thought that, because the OP's cross hole is blind, at least we would be able to go to a projected view where the hole is shown "end on", and assign a center mark to the interior end face of the hole, given it is a true circle ...
but again we are thwarted, because Onshape's centermark tool cannot "see" the circular edge to pick it, hidden as it is behind the silhouette edge where the cylindrical hole breaks out of the shaft. ("Select Other" is not available in Drawings)
I tested this by adding draft to the hole (selecting the end face as the Neutral Plane) and found I could then use the centermark tool, and dimension the hole diameter in that view.
Moreover, I could dimension from the centermark to the end of the shaft, something I had not otherwise managed.
So in the OP's case, a (nasty kludge) workaround would be to do that (as shown below), then reduce the draft angle to a very small angle indeed so the lines "look" right.
But it relies on happenstance: the hole happens to be blind, and the end face happens to lie on the midplane of the shaft.
I'm frankly embarrassed to suggest this kludge, but not so much on my own behalf.
I was thinking of turning this search for a workaround into another "Random Challenge" , but Onshape's drawing environment is currently so barren of capabilities to work with, it seemed to me it would be no fun at all, and probably attract very little by way of participation.
For instance: we can sketch lines, but we cannot
a) constrain them, even to be orthogonal (H or V), let alone to view geometry.
b) snap them to view geometry
b) dimension from them to view geometry (even driven dimns, let alone driving)
Nor can we access sketches from the part in order to snap dimensions to them.
Using the drawing package as it now stands is a bit like trying to fix a mechanical watch wearing boxing gloves.
I respectfully disagree that the 2D problem is difficult to overcome. In fact, if machining is ever added to Onshape, as is being discussed, it will need to be addressed in order to do XY roughing of 3D shapes.
I was involved with this in decades past. One easy solution is to just put in a menu option to process splines into arcs when projected onto a plane (and 3D arcs as well for machining purposes). You put in a tolerance factor and, internally, generate points along the spline to generate circles from sequential sets of 3 points until the radii and center points diverge. The arc segment can be highlighted for the purpose of dimensioning both the radius and centerpoint.
This goes back to even the oldest CAM systems such as MDSI, where "tabcyls" (tabulated cylinders) were converted to arc machining in post processors. I have also used it when capturing streaming data from Bridgeport tracer mills to generate 2D CAD drawings from broken parts that were crazy glued together, then passed to CAM systems. And that was in the mid 1980's.
I think we might be talking about different types of difficulties.
You seem to be talking about converting 3D geometry into 2D toolpaths.
I'm talking about 2D Drawings, as a way of conveying 3D geometry to a human machine operator (who is running a manual machine tool, or programming an automatic one).
I think the difficulties which trip up(or slow down) drawing module development, in cases where 2D linework is automatically generated from 3D models, are mainly around providing clever-enough interface tools (and crafting them to be usable enough) for fast creation of unambiguous drawings which are easily human-interpreted.
And I think this is because the 'language' of 2D drawings is (of necessity) sometimes abstract rather than literal.
It seems to me we only realise this when we try to automate their creation, because abstraction is essential to human intelligence, built into our DNA, and hence somewhat invisible to us.
Artificial intelligence, in striking contrast, struggles with abstractions, and often does not live up to the second part of its name.
If you have a 3D spline, you have a formula attached to it, perhaps a cubic descriptor. You can take that spline and generate points along its length. Those are 3D points. Select an plane. If it's a simple basic plane such as top, front or side, just drop one of the three arguments. For example, for the top plane, drop the Z descriptor. You are left with a 2D array of points.
Now do a simple loop, checking that each point is at least a minimum distance from its neighbor (nearest neighbor algorithm with mindist culling system). That would get rid of a steep vertical slope in this instance, and in any instance will deal with points too close together for the generation of credible circles/arcs.
So now you have a 2D array of contiguous points, on which you iterate for persistent arcs and lines. If you put this in the modelling workspace for selecting centerpoints of intersecting cylinders/holes/solids/surfaces and in the drafting workspace for dimensioning, the problem goes away.
I've been involved with CAD, CAM, FEA, and coordinate measuring machine software. I've found that one department is often averse to using algorithms utilized by other departments because they were written for a different market. But an algorithm is just an algorithm and has no allegiance.
I don't personally imagine it's the automatic geometry creation issues you are focussing on, or in fact any issues solvable by geometry and topology boffins and geeks, which tend to hold up the development of usable 2D drawing packages for bolting onto 3D modellers.
But I do think your points are interesting, because it seems to me that most developers DO focus - at least in the early phases - on issues of the sort you describe, to the exclusion of issues which are of more concern to the user, which I described above as "providing clever-enough interface tools (and crafting them to be usable enough) for fast creation of unambiguous drawings which are easily human-interpreted."
Taking your example of exposing underlying geometry to the user for selection: in a complex model, where the underlying geometry competes for the user's attention, it seems to me the real challenge is not how to expose it, but how much to expose it, and how to allow the user to modulate (and navigate through) what is exposed.
But that (it seems to me) is still easy compared to issues like non-literal portrayal, in the interests of clarity and interpretation.
And if Onshape seem to be dragging the chain on little things which (as you point out) are easily soluble, it may be a reflection of how much resource it takes to try to solve the big things.
I might be behind the times, and that might be because I was once a detail draftsman on a physical drafting board. Today I think of a CAD (or CADD) system as a tool, used by a designer, that provides numerical data used to create a drawing, run a machine (additive or material removal), drive an FEM system, do simulations, generate bills of material, automate tool design from the piece part, give animation and simulated photo output, and all the other functions pertaining to design and manufacturing.
Blueprints are on the decline in many disciplines, and in my industry are used mostly to provide tolerances for manufacture. Even if the drafting portion of the system is flawless, it is useless to me if it won't eventually have hooks in it to pull the data out and feed data in.
I'm new with Onshape, but I got involved because I read somewhere that we could get data in an out via python. The data must be accessible in any modern system for the system to be of use, at least in my opinion.
Certainly the user interface is very important, but to the power user functionality is key if the UI allows one to get started within an acceptable timeframe.
I guess what I'm trying to say is that to me, a CAD system is a tool for designing a part or assembly, not a drawing. The drawing is just one of its useful outputs. Concentrating on auto-dimensioning or other routines that save a little work instead of investing resources to increase functionality would be a bad choice, I think.
I guess, because the difficult you originally posted about was producing a Drawing, I thought that was what we were talking about.
It is certainly what I was talking about.
I mentioned the dimensioning issue because I thought I just didn't understand the user interface, and was asking for help in learning how to use the system to dimension the hole.
I neglected to mention that I couldn't even pick the hole in the solid model. The way I put a point there in my later sketch (hoping to be able to dimension to that point instead of the circle center, which didn't work) was to just manually put in a point a certain distance from the end of the shaft. I figured that once I discovered how to pick the hole for dimensioning, I could then pick it for modelling selection.
Then, from the answers, I discovered that the dimensioning issue is just part of a larger issue of not recognizing that what Onshape sees as a spline, is a circle from above. It's basic to the data structure.
If it were just a dimensioning issue, then adding the ability to add sketch entities to the dimensioning workspace would solve that easily.
I get the feeling I upset you. Let me give you some background, if it helps. I have been using CAD since the early 1980's, have been involved in the industry for a long time. I wrote software and trained people in CAD both in industry and universities. I've spec'd out systems as well. Now, in my old age, I am a user.
I have used many CAD systems, including SWorks, Inventor, many others. I believe cloud CAD is a fantastic idea and am migrating my system to that end. I am updating my hardware, since my old twin Xeon system doesn't work with Onshape. So I am committed to that end.
The only two systems that I found with any user base are Onshape and F360. I'm not a fan of Adesk (though, or maybe because, I sold their software for a while). So I'm trying to get a handle on what Onshape will be able to do in the future.
I could have modelled the shaft in any CAD system. The only way to learn about a system is to use it. I have started to do so, and am finding its limitations. I thought that it must be my limited understanding of Onshape's user interface, so I asked for assistance in that regard.
My concern from this thread is whether such limitations will be addressed, or whether the people at Onshape feel that it is too much bother.
Drawing is the least complete part of Onshape and as Andrew has alluded to, the rate at which features and improvements are being added into drawings seems to be slower than the modelling tools. It's a little frustrating when trying to complete a drawing and the tools just aren't there, even for work arounds. The speed of drawings has also been an issue and I believe optimisation is still taking place to ensure the speed is acceptable. I am hoping that once the speed and stability of drawings is fully sorted we will see these key features roll in very quickly. If your are missing bits as you discovered I'd suggest filing a ticket to help Onshape prioritise development, the more tickets and demand the more likely we will see the improvement.
Twitter: @onshapetricks & @babart1977