Had Another Go

[Charles LamontParticipant @charleslamont71117]

Nigel says he can’t view the image I posted above at #780279. Others having trouble?

[JasonBModerator @jasonb]

It was there earlier then went, probably be back tomorrow

[Charles Lamont]

On Charles Lamont Said:

Nigel says he can’t view the image I posted above at #780279. Others having trouble?

Excellent Charles! I’ve been trying to catch an example of this bug for ages, and yours let me capture the evidence.  I’ve reported it thus:

Forum malfunctioning with posts, images and maybe messages not appearing until later, and maybe randomly disappearing/appearing. Most common with images in new posts. Text appears but the image doesn’t (defaults to alt text). Later the image does appear. In example provided the image URL gives a 404 error. Also noticed but difficult to catch are entire posts ‘delayed in transit’. Snot and feathers fly because people reply to what appears to be the latest post, and it isn’t. Please fix: when the delayed post finally appears it’s inserted into the stream, and looks as if it’s always been there. Database or lazyloader problem?

Screenshot provided showing the broken URL, which causes a 404 Error:

The image should be here and isn’t.

Intermittent bugs are the worst.  Oh, and as a reward for reporting the bug I was logged out. Pretty sure that’s yet another intermittent bug.

Cheers

Dave

[Charles LamontParticipant @charleslamont71117]

Why does the .png I uploaded need to be converted to a .jpg?

[Nigel Graham 2Participant @nigelgraham2]

A few hours and a few score wretched constraints errors and bits drifting out of position after posting that….

It should look something like this, if we ignore that the base-plate has slipped back to its favourite spot slightly in from the edge of the wall, and the three main bearing holders (of temporary / unfinished form) are not properly located:

In practice the two outer bearings need be in housings on the end-covers, otherwise it would not be possible to assemble or service the engine. Also the cylinders will be slightly higher. The “front” of this elevation (actually facing the rear of the lorry) has quite a distinctive shape I need somehow replicate. It’s not a plain rectangular box from top to bottom.

 

[Charles Lamont]

On Charles Lamont Said:

Why does the .png I uploaded need to be converted to a .jpg?

Pass!

With my Software Engineer hat on, probably a simplification.  The forum resizes images to make them web friendly, so:

1. Any allowed image format that’s not a JPG is converted to JPG.  Easy to do and avoids a lot of ifs and buts later.
2. All images are resized as JPG. This is an efficient compressed format, saving space on the server, reducing network load, and compatible with all browsers.

I’m guessing!  Moderators don’t have access to the forum’s innards.

Dave

 

[JasonBModerator @jasonb]

It’s showing now Charles just like I expected.

I can also see that it is the usual (unexplained) problem of an image placed as an edit tends to cause the problem rather than one placed as part of the original post. Had you left your original image I expect it would have stayed visible.

[SillyOldDuffer]

On SillyOldDuffer Said:

On Charles Lamont Said:

Why does the .png I uploaded need to be converted to a .jpg?

Pass!

With my Software Engineer hat on, probably a simplification.  The forum resizes images to make them web friendly, so:

1. Any allowed image format that’s not a JPG is converted to JPG.  Easy to do and avoids a lot of ifs and buts later.
2. All images are resized as JPG. This is an efficient compressed format, saving space on the server, reducing network load, and compatible with all browsers.

I’m guessing!  Moderators don’t have access to the forum’s innards.

Dave

 

As a mere user … but one who frequently posts large images in various formats … I feel sure that Dave’s analysis is correct.

The upload of images in various formats is explicitly encouraged, but ‘the machine’ then processes them to suit its own best interests.

MichaelG.

[Nigel Graham 2Participant @nigelgraham2]

Ahem!

Back nearer the topic, while you were all discussing image formats I know nowt about I was footling around with own image, below. So “succeeded” in throwing together something vaguely nearer to what my model engine is meant to look like, with various simplifications so I could draw it.

The big lower “windows” will have what the prototype’s advertising called “quickly detachable covers” – held on with lots of wing-nuts.

Measuring shows the whole unit just in the height limit of 16 inches overall, give or take a nut or two.

No it won’t twirl elegantly round if you try dragging one of the crank-webs (on the model not this screen-shot). Well, it will, but likely push the two eccentric sheaves and one strap I have included, off sideways… As it did me, and on the ceiling too, after the umpteenth constraint failure, placement-tool mutiny and Part drift.

For simplicity I gave the CAD model’s eccentrics “virtual keys”, only to find the sheaves impossible to align and constrain properly with the shaft and its keyways. Any attempt says “over-constrained”. The real engine will be fine with just shallow drilled pits for grub-screw points.

I’ve not yet drawn the rest of the valve-gear parts I will not be able to add to this assembly anyway.

So this is Edition Two of a very rough, provisional model: several hours’ struggle, very inefficiently drawn, full of CAD errors; but only about fifteen parts an expert would likely assemble in fifteen minutes with no mistakes.

At least I can take sufficient measurements to assess just where the other bits will go, which is its purpose!

[Oh. Oh dear! About to press ‘Submit’, I spotted something odd about the eccentric strap. Went back to the original Assembly, enlarged it and yes – the strap has slid off the sheave and partially melted itself into the crankshaft. Such is life with CAD…]

[JasonBModerator @jasonb]

You have come a long way in the last month, I think using it everyday ha sbeen a big part of that.

 

Offer still stands if you want an assembly to play with and get the parts moving. If you send all the part files and the assembly file I may only need to do a few alterations to your assembly rather than put teh lot together

[Nigel Graham 2Participant @nigelgraham2]

Thankyou Jason.

The faults in that assembly are most likely wrong constraints, because I don’t know how to choose what should be constrained where, and which way round, in any complex model, not just that engine. So cannot ensure any constraint will work, or that later additions won’t pull existing parts out of line, or that any animation of moving parts will not break the model.

(… way round… Does the order in which objects or planes A and B are named on the form, decide if A moves to B, or B moves to A?)

 

I still struggle to align Parts to faciliate Assembly. They come in upside-down or the wrong way round.

I don’t know the proper conditions for that “precise placement” tool to work. I find even that roughly dragging it around by the cursor and Triad very awkward.

I don’t even know how to plan any Assembly of more than three or four Parts so I import them in the correct order for it.

.

Yes, frequent use helps but I should have advanced far more than I have, long ago. I think I have reached a limit, or “plateau”.

I have long thought the limit exists, but heard of the plateau effect only recently when a speaker on the radio talked of pianists unable to develop their skill above some unexpected level.

The context there was rather odd: a sort of mechanised training-glove a Japanese engineer has invented to help pianists increase their dexterity. I doubt ability to play dazzling cascades of allegro demi-semi-quaver arpeggi is very relateable to ability to depict fairly complicated machines on a computer, but the musician’s “plateau” concept certainly is.

Further, you are not on a plateau if you can go higher. You are still climbing the hill-side. Once on the plateau, the only ways on are round in level circles, and back down the hill.

……

(Oh, and no I can’t play the piano… The drums, once, yes – though I never progressed to the arcane arpeggio equivalents called “Triple-Flam Paradiddles”, introduced late in Buddy Rich’s tutorial book! )

[David JuppParticipant @davidjupp51506]

The order MIGHT decide which moves, but any other constraints that already act on either member will influence things.  Start from a part which is already fixed in place, then constrain others to it.

Using sub-assemblies is often useful, as it reduces the amount of constraints at any one level – hence reducing likelihood of conflicts.  You only need to make a sub-assembly ‘flexible’ in an assembly if needed to allow a motion.

 

[Nigel Graham 2Participant @nigelgraham2]

Thank you David.

Constraints? I can’t even draw a single part without it going haywire.

This link (below) is copied from a published design, though I have thickened it slightly to give a little more bearing surface for the die-block that works in it.

A pair of small brackets is fitted centrally to that wide flat area on the right, on each side of the plate, with three small rivets in a triangular pattern. Their dimensions are on the separate drawing for the brackets, and not at all easy to determine.

The original drawing is inordinately hard to decipher thanks to a mixture of odd vulgar-fractions in chain dimensions, and several radii of which only one is annotated. A right so-and-so whether you use it directly in the workshop or try to translate it to CAD with decimals fractions. To place the two holes you even need know how the whole caboodle goes together, to be sure of the way they are dimensioned.

However, I managed to re-create the main plate, though the two back corners need fattening up a bit.

 

Then I tried to add the rivet holes, needing more calculations of the original print because its draughtsman had not dimensioned it fully. Two things happened:

1) Despite great care to start with making it all axial so the difficult geometry would work, it has tilted over slightly and moved vertically downwards, so now lies randomly – useless.

2) The original sketch is mentioned if I edit the extrusion; but has fallen off the Design Explorer so I cannot view it to edit it – useless.

 

Even if I could see it, the outline has moved randomly, so plotting the rivet-holes is now impossible. Those brackets hold pins about which the link as a whole oscillates, so if not located accurately the entire link is useless. From a CAD point of view, it’s now impossible to locate them because it has lost all its references and the original sketch.

So now what the heck have I done wrong? I can’t restore the link to its original orientation, so will have to delete the entire file and try again – another hour or two wasted.

 

In the image, the two outer corners of the slot might not lie on the axis but if they don’t, they should be the same distance from it.

The whole thing should lie equally about the axes, as I started it, on which I placed the centre for the 5 or 6 circles needed for the construction geometry and parts of the outlines.

 

[DiogenesParticipant @diogenes]

Has the sketch gone or is it just hidden – if you toggle this does it come back..?

[JasonBModerator @jasonb]

What you have done wrong is attempting to add the three holes to the original sketch.

Far better to have the basic shape as the initial sketch and then do the holes as a separate sketch.

To bring the sketch back onto the list click the “>” next to Extrusion1 and that will expand it to show the sketch, expect you clicked the downward pointing arrow next to it by mistake which hid the sketch etc.

There is another thread where we played about with this link before.

Don’t delete it send to me or David. You did not have the initial sketch constrained enough otherwise it would not have been able to move.

[JasonBModerator @jasonb]

I’ve found the old thread here

Even if you can’t reposition the original sketch the three holes can be positioned relative to the two larger holes so send the file and I’ll show that option a swell as getting the sketch back to where it should be.

[Nigel Graham 2Participant @nigelgraham2]

Thank you.

Oh, I tried everything to restore the original sketch, and nothing worked. Nor could I find out what what I’d done wrong, but the original drawing is poorly dimensioned and I might have let that confuse me somewhere.

I deleted the entire file, started again, now putting the little rivet holes in at first. This time it worked, and I went on to creat the little brackets. I had to modify them when I found I’d made a dimension mistake, and I am not sure it’s right now. On the print, its key dimension is not on its own drawing but derived by subtracting an offset from a radius on the link drawing, then cross-referencing the two drawings! The bracket is annotated to say its back face is flush with the link face, but that is not dimensioned from anything, and the rivet pattern is dimensioned internally but not from anywhere else.

So if you find CAD difficult when trying to translate an original drawing, that being dimensioned in complicated, unclear ways doesn’t exactly help.

I managed to assemble the three parts, but to verify the dimensions, presumably have to derive a dimensioned drawing from it.

The larger holes through the two brackets take pivot pins. Their centre-line has to be set back from the centre-line of the slot in the link, by an important dimension among those least clearly defined on the original print. I have already corrected it once and am still not sure it is right.  The 1/2″ dimension for it from the back of the link is marked as “nominal” and is from a surface not dimensioned to anything, not even by corner-rounding radii!

[David JuppParticipant @davidjupp51506]

And besides, does it really matter if the part is displaced from the default planes/axes?

But yes as Jason says – send the file to one of us.  It’s simple enough to fix – either with sketch constraints or with the sketch Move tool.  I can even do a quick screenshare to show you how.

Personally I wouldn’t put so much into a single sketch.  Complex sketches can be problematic to edit.  Use one sketch for the outline (even that I’d simplify by not including the rounded corners – add those later with fillets), another sketch for the curved slot (perhaps use offset tool from the curved edge), and another sketch for the holes.

That approach will give faster response, and if you need to (say) edit hole size or position there’s no risk of disrupting the part outline.

Ideally all your sketches should be ‘fully defined’ – so that they can’t be accidentally moved.  The Degree of Freedom Colours display helps you with this.

[David JuppParticipant @davidjupp51506]

I managed to assemble the three parts, but to verify the dimensions, presumably have to derive a dimensioned drawing from it.

Well that’s one option – but there is a measurement tool in the 3D workspace too.

[David JuppParticipant @davidjupp51506]

So if you find CAD difficult when trying to translate an original drawing, that being dimensioned in complicated, unclear ways doesn’t exactly help.

It doesn’t – BUT with CAD you can try it, check it, adjust it (and repeat until it makes sense).

[Nigel Graham 2Participant @nigelgraham2]

The image above is wrong. As I feared.

It took a further two attempts to set those pin holes in the brackets correctly: each time go back to the Part, edit that, save it and resave the assembly, create a drawing, set it to inches instead of mm, dimension the two important measurements in the drawing (not very easy)…

All very frustrating and disheartening though better than making three pairs of expansion-links until one is right.

The remaining parts should be simpler although there is at least one ambiguity in the original drawing.

I do wonder how many physical models have been started and eventually abandoned because the unfortunate, self-taught builders just could not overcome a mixture of their own mistakes and mistakes or ambiguities in the published plans.

Not that CAD stops you making mistakes in your own CAD models, provided they are permissible to the software.

[Nigel Graham 2]

On Nigel Graham 2 Said:

It took a further two attempts to set those pin holes in the brackets correctly: each time go back to the Part, edit that, save it and resave the assembly, create a drawing, set it to inches instead of mm, dimension the two important measurements in the drawing (not very easy)…

 

Nigel – you don’t need to do all that each time.  Just edit the part.  When you re-open the assembly it will have updated.  When you re-open the original drawing, it will announce that re-projection of views is necessary – OK that and it will update.

There’s absolutely no need for all those extra steps you mention.

[Mark EasingwoodParticipant @markeasingwood33578]

Nigel, you are making good progress with Alibre.

I would suggest, that, what David outlines above is one of HUGE advantages of CAD, over pencil & paper!

Mark.

[JasonBModerator @jasonb]

You should also not need to keep changing between Metric and Imperial unless you are working in both.

If you open a new part, then click “File” top left. Come down to “properties” and click that it will bring up the box where you can change units.

Change from metric to imperial and then the important part, click the make default box bottom left so a tick appears in the box. Then click apply and and then click close. From then on all new parts will be in imperial unless you change it.

Same with drawings set it to imperial and make default then it will remember that. Click Dimension Style, alter to imperial, tick box and apply

[David JuppParticipant @davidjupp51506]

For 2D drawings, setting the default units as Jason suggests only works for Blank Sheets.

The 2D drawing templates have their own settings – which any drawing made using the template inherits.

 

You can add extra dimension styles – and switch between them, even in the same drawing.

 

[Author]

Posts