How Am I Expected To Know What This Means? (Alibre Atom)

[JasonBModerator @jasonb]

20 constraint icons showing in the sketch I posted so maybe I’m doing something right😊😊😊 would be more if I had not purposely made the sketch with the tiny error

Yes that is the way I tend to do things place the circle approx where I want it and then use a concentric constraint to line it up with the hole already in the link, snap will put my dividing line central. Draw a rectangle and then size it and make it symmetrical or align a side with an existing feature.

[NealebParticipant @nealeb]

Hope you don’t take my comments amiss, Jason – not your sketches I was thinking about! And I do see on closer inspection that at least one of Nigel’s sketches include a few constraints – although not as many as I would expect. Of course, you can’t tell whether or not they were explicitly added or automatically added by Alibre.

The real point is about the general approach, which really needs to play to the strengths of modern software and not apply traditional draughting techniques which are designed to achieve a very different end. Every screen used with 3D CAD needs to have “a sketch is not an engineering drawing” engraved across it!

I shall now retire to my own CAD system and continue using old 2D drawings to construct a 3D model of my loco under construction in order to find more errors – which would not have occurred had Don Young had access to modern software!

[Nigel Graham 2Participant @nigelgraham2]

Nealeb –

No, I am not trying to apply manual draughting techniques to CAD! I know about that trap. However…

Yes, I am trying to produce engineering-drawings. You need the elevation drawings the software derive from the 3D model, so that model is an indirect route to the same practical end.

The caveat I do know, is that a 3D CAD model of a complete assembly helps in assessing the whole design more easily than from orthographic drawings.

Except.. I can’t, but that’s not a fault of the software, just my low ability. I tried an Alibre assembly of my model engine: nothing was quite where it should be and it all fell apart when I tried to rotate its crankshaft.

 

I have to use TurboCAD Deluxe’s orthographic option if I want an elaborate assembly drawing.

This does need adapting conventional draughting methods, as you create conventional elevations in much the same way. It also allows complex geometrical constructions, subject to you knowing their drawing techniques, of course. I have old text-books explaining them.

3D software may well generate those for you, irrespective of the make, but that’s beyond my level. I cannot use TurboCAD in 3D for more than very basic, single items.

So its 2D mode TurboCAD is a direct analogy of the drawing-board, though far more accurate.

Whereas TurboCAD in 3D, like Alibre, SolidEdge, etc. lets you create accurate “pictures” then it can plot the consequent elevations for you. I think: I don’t understand its “Viewport” system properly.

 

Various people have suggested SomeOtherCAD instead, but Alibre and TurboCAD have already cost me much brass and time, so I do not want another expensive fight and same risk of failure!

I do what I can with what I have.

Hence I can use TurboCAD Deluxe in 2D mode enough for moderately difficult assemblies and parts; Alibre Atom for fairly simple parts.

What I don’t have is what everyone else has: ability to learn advanced CAD including 3D assemblies, in any package.

.

Jason –

There are some tools I never have much luck with.

That symmetry constraint is one, fillet is another.

While for Assemblies, well all of it really.

 

 

[Nigel Graham 2]

On Nigel Graham 2 Said:

 

Except.. I can’t, but that’s not a fault of the software, just my low ability. I tried an Alibre assembly of my model engine: nothing was quite where it should be and it all fell apart when I tried to rotate its crankshaft.

.

Jason –

There are some tools I never have much luck with.

That symmetry constraint is one, fillet is another.

 

 

 

Falling apart when you try to rotate the crank is likely to indicate that the parts in the assembly are not fully constrained but just placed by eye. Therefore, they move in more than one direction when all you want to do is drag the crankpin around with the mouse.

Symmetry is one I use a lot as well as fillet so worth spending some time trying your luck. The one thing I find with symmetry is it can sometimes need the mouse hovering over a differnt part of the first axis line in orfer to get it to change colour and be selected.

With paper and pen you would take your rule and mark a distance then draw the line through that point. With CAD it is really the opposite. Place the line often letting it snap vertical or horizontal. After that to position it with either a dimension or constraints and then give it a length or trim it to other features. This also helps to ensure the sketch is fully constrained and you have not just placed something buy eye. Constraining also helps when you come to make changes as if features are correctly constrained anything related to that feature will adjust itself.

Taking that link rod if the semicircle were not concentrically constrained to the hole in the rod and you subsequently altered the length of the rod then you would end up cutting away more of less of the rod and not rounding the end as it should be.

 

Taking Nigel’s earlier image I would not expect to see much more in the way of constraints. Alibre won’t show the constraints that were used to construct other features that made up the square ended rod. It has snapped the 0.375″ circle to the dashed green axis line, the green double ended arrows indicate the diagonal line is vertical and the other three lines that make up the shape are not critical so can be unconstrained. What I can’t see is the concentric constraint symbol to tie the whole thing left to right although the inner hole has changed to green which can indicate it is or was being used as a reference. If it is not concentric then Nigels 2D sketch could be too far to the right and that is why there is a flat spot on the end of the rod rather than the fillets meeting.

[Nigel Graham 2Participant @nigelgraham2]

Thankyou for explaining it, Jason.

I didn’t even think about constraining the two circles to each other – didn’t occur to me, or I didn’t realise the possibility. I set their centres by dimensions from the flat surface at the far end.

The hole and arc are where they are supposed to be. The flat is more likely from making the initial outline slightly undersize, perhaps as I think you suggested previously by under-rounding the dimension. The physicall making wouldl be cutting steel back to the intended size so should not create a flat in the real thing.

The critical position of course, is that of the pin-hole; more specifically as this is a Stephenson’s Link Motion eccentric-rod its equality in all four rods. A small length error is unlikely to matter, within reason, as long as all four rods are of equal centres. Besides, I cannot calculate the theoretical centres to Churchward or Gresley standards, before typing them into CAD dimension boxes! I am adapting a design by K.N. Harris after LBSC, as being close to the nominally-scaled engine I am trying to build.

Since then I have wondered if round rods with a clevis screwed on the top end and a foot screwed on the other would be easier to make, and use less material! Or silver-brazed then finish-machined. (The former version would need opposing-handed threads to set the distance and orientation.) Some full-size engines did use thread-ended rods.

At least by drawing items first, however well or badly, allows me to assess possible alternative ways to make them.

[Author]

Posts