Planes and Moves/Rotations in Alibre Assemblies

[David JuppParticipant @davidjupp51506]

Nigel,

In your image of the crankshaft assembly, the crankshaft itself is well away from any axis of the assembly workspace.

If you haven’t fixed it, it will be free to move.  This can make constraining other parts to it more of a challenge.

If you’ve locked it, it won’t be free to rotate and you will run into issues if you combine this assembly into another assembly and want it to move realistically.

I would suggest constrain the crankshaft;

• co-axial with one of the reference axes of the assembly (use either an axis of the part, or any cylindrical face of one of the main bearing journals or crank nose or tail).
• Axially to an assembly reference plane (use either a plane from the part, or any suitable flat face).

[Nick WheelerParticipant @nickwheeler]

Your one hell of a struggle is self-inflicted:

Attach the small ends to the conrods and join each complete part to the crankshaft.

Alibre must have a way of putting joint origins(a Fusion 360 term, so it’s probably called something else) in the centre of the big end hole and the crank pin diameters. You shouldn’t need to work out where those are, as the computer can do it far better than you. It won’t be the only way of easily making these joints….

Nor should you need to plot where the other ends of the eccentrics or conrods go, because you’re going to use a similar joint for where their pins operate the semi-reciprocating side fumbling dingle arm or the piston’s gudgeon pin. Attach that using the inherent geometry of the parts, and the whole mechanism will shift into place without any further ‘help’ from you. It does assume that your parts are actually capable of working, but that’s also true of assembling them when you’ve physically made them. Far better to find out before you do that.

If you modelled the parts using defined parameters for measurements between the pin centres or the eccentric offsets(crank throws, cylinder deck heights, whatever)you can easily adjust out any issues that you designed in just by changing some numbers.

To really use the power of CAD, you wouldn’t model the conrods or eccentric straps until you have the pistons and crankshaft working in their bores/mechanisms.

[Nigel Graham 2Participant @nigelgraham2]

[Odd. I opened the site from the link in the e-post notification, and found I was already logged in. Then remembered that by following a link within it last night I lost contact with the ME site so closed it by default. Since the computer was off all night I’d imagined it would have been logged out automatically.]

To answer though….

I saw the Assembly planes are not those on which you start the initial sketches and model, and that does not exactly help matter!

I had built the crankshaft up from a single, mid-planar, extruded bar along the Y axis (0,Y,0).

Created the webs and crank-pins, and added them, then managed to find how to re-visit the assembly so far in sketch mode to remove the bits of the original bar.

.

In my previous attempts, laboriously built up in a different way, I discovered on assembly only part of the shaft was anchored and trying to put other bits on would break it up.

So this time I saved the new shaft as above as its own file as if a single part. Then started a new assembly with that lot anchored, to start putting the other bits on. There are two pairs of larger-diameters in there, and I made those as individual Parts called “sleeves”.

In all I made three assemblies with the previous anchored as a single “part” for adding more details. The final bits I added were those Impressionist eccentric-rods and so they are not anchored.

It did indeed lie well away from the indicated planes, and this seems to happen automatically. It opens with its length lying vertically on the screen, with me looking down onto it, and what the assembly view-cube calls the “top” and “bottom” are the end views not “front / back” or “left /right” .

So I was not sure if rotating anything about the Y-axis would turn it round the shaft or remove it. Luckily it did move as I wanted.  I did try to constrain the assemblies’ own reference planes to the main ones but could not make that work.

.

You describe the problem of “locking” it. Is that the effect of using “anchor”?

I had not even considered trying to plot movements of machine parts. I know it’s possible but far beyond my level now. I’m still trying to understand how to hold things still and in the right places so I can fit others to them!

For practical applications, I’d find it simpler and more certain to plot the loci and end-points manually, on prints. It would not need whole copies of the parts, just enough to show limits of movement.  For example, for the space needed by a connecting-rod big-end, its corners’ and studs’ loci are just concentric circles on the end view axis.

[David JuppParticipant @davidjupp51506]

Yes – Anchor in assemblies ‘locks’ the position of the part completely and also prevents rotation.

Any part first appears in the assembly wherever your cursor is on calling the part – it’s up to you to choose the location when you click to place the part, and with any assembly constraints that you later apply.

[Nigel Graham 2Participant @nigelgraham2]

Nick –

Self-inflicted? 🙂 I do know expert users would draw that lot far more efficiently and easily than I can, and from scratch not existing parts!

Most of those components I’ve depicted are simplified from parts I’ve already made, so I just measured them. The real connecting-rods are rather more elegant than shown but too hard to draw fully.

The eccentric strap and rod is shown as a single part for drawing simplicity; though it’s made me realise I could make the real parts that way from steel plate, fitted with one-piece liners and assembled from the side of the eccentrics, with separate cheek plates to hold them on the sheaves.

Your Fusion’s “Joint Origin” presumably corresponds to Alibre’s “Concentric Constraint”, which asks you to select the mating curved surfaces rather than the centres to put a cylinder on another. At least, in how you use it and how it seems to “grow”, whatever is happening hidden behind the screen.

From that, I then discovered I could place the sloping connecting-rod’s small end on the vertical plane by another Constraint. That made the rod swing as it should, round the crank-pin.

.

That drawing above, and all its preparatory ones, are an “exercise”. They borrow from a real project but live in their own folder called that. If I am to use Alibre to help me design (it can’t itself “design”) the real engine parts are in a separate directory, and its drawings also detailed as far as I can.

The engine is enclosed and already set, more or less, in itself and in its place within a vehicle layout I think unique among steam-wagons. Much of the basic layout already exists in orthogonal TurboCAD drawings I started a long time ago; but these are many-part general-arrangements with poorly-sorted layers and peculiar “work-plane” / co-ordinate errors. So I cannot transfer them to Alibre even by standard, transferable file-types both makes can use.

So the order in which I draw the engine parts and sub-assemblies does not greatly matter here, except for those between the cylinders and cross-heads. It certainly won’t make any difference to the problems of accommodating the transmission from engine to axle, the feed-pump and brake-gear. With no known surviving drawings, I have no idea how the original vehicle’s designers solved those problems in 1908, but they did – and on paper on elm drawing-boards with loose squares.

[Nigel Graham 2Participant @nigelgraham2]

Thankyou.

Complex? Yes – I think I’ll try something simpler for now….. That assembly was a bit much.

[Nigel Graham 2Participant @nigelgraham2]

Like this….

Steering-wheel. 4.75″ dia, which seemed about what the photograph gives and scales to 14.25″ full-size so probably not far off. The small hole in the rim is for a vertical handle.

The central bore is nominal so far. I’ve yet to work out how to fit it to the shaft.

It was only after I laboriously rounded to rim by an extruded cut then the inside edges by “Fillet” I realised I could have used the Fillet tool for the rim as well!

[JasonBModerator @jasonb]

Yes several ways to do things, another would have been to draw a cross section of the rim and revolved that as a solid. That way you would not have the problem of needing a smaller radius on the inner edges to avoid cutting through the spokes.

Assembly to shaft is much like you have done with the crankshaft and it’s parts: Concentric constraint of shaft to hole and another constraint of the underside of the hub’s face to the end face of the larger dia of the shaft

[Nigel Graham 2]

On Nigel Graham 2 Said:

Thankyou.

Complex? Yes – I think I’ll try something simpler for now….. That assembly was a bit much.

Much truth in Nick’s comment “Your one hell of a struggle is self-inflicted“.  Trying to run before one can walk is always painful, and moving to model the crankshaft assembly immediately after trouble with the block was several steps too far.   I’m impressed by how well Nigel is doing considering how difficult many of his ‘simple’ tasks are.

How best to learn is personal, but the usual method is to start with the basics and then advance step by step.  When a step fails, it means that an earlier lesson wasn’t quite understood, and it’s necessary to go back and fill the gap.  Frustrating, but I don’t know of a better alternative.  Leaping ahead doesn’t work for me.

The steering wheel is a good learning project, and I recommend exploring it thoroughly.   There are a number of ways of modelling it as a single part, all with pros and cons.   Worth experimenting with a steering wheel made as a:

• Single part, from a flat disc, spokes cut out, edges filleted
• Single part, from disc, produced by rotating an extrusion. then cutting out spokes with a pattern
• Assembly of a single part rim, with ‘n’ spokes, and a single part hub.

Then address how the wheel is fixed to its shaft in various ways:

• Keyway
• Splines
• Taper with bolt
• Flange on shaft with bolts

All simple to do, but learning how for the first time requires a fair bit of effort.    But once mastered, they’re all are useful for making other parts.   Just as important as individual methods is learning when to apply them or not.  I’d probably model a single part steering wheel by rotating an extrusion because most of the shaping is done by the sketch.    Not the law though – other methods suit different circumstances.   For example, filleting the rim with an rotated cut is a valid technique and Nigel learning it is good because the rotated cut method can do complex profiles whenever fancier than a straight chamfer or round fillet is required.

Even though I’m fairly CAD proficient, I keep an eye out for common objects I don’t know how to model for more learning.  When I have time, this cafetière is next:

 

I don’t know how to put a spout on the jug, or create the curved wire-mesh filter, or the coil-spring that guides the piston.  Although the handle on this particular example happens to be easy, others have complex finger grips. Though I’m sure it can all be done, that’s 4 things on a basic coffee pot that are beyond my CAD skills.   Though I shall never be a fully qualified CAD jockey, I can do most of what I need, which is is OK until I get stuck again!

Dave

 

[Nigel Graham 2Participant @nigelgraham2]

Thankyou Dave!

I didn’t draw that wheel right through in one go.

It took a few attempts to get it right.

My first mistake was laboriously plotting all four cut-outs, which meant chasing very many misconnections between lines and arcs. Then I spotted the sketch fillet-tool and worked out how to use that for the corner radii.

The spokes now “worked” but looked too wide so I remade them, this time using the radial pattern tool to copy just one plotted hole.

Making the “form-tool” for the rim radius needed a few attempts too. The first left a razor-edge “burr” round the wheel, so I made the second generating circle diameter very slightly larger than the wheel thickness. Then had a minute fragment of line from somewhere in the construction, that proved very difficult to track down and delete.  I’d also extruded the wheel from the flat plane, not through it, making this construction harder.

.

That MEW introductory series, drawing a scribing-block, did have two extruded-cut operations! One was a taper to fit the column to the base, and that was not very easy; the other was even harder, the thread on the scriber-clamp.

Some while later I tried to draw an M26 bolt and nut, from measuring said items, using the same methods. I think I managed it, but cutting the thread was certainly no quick and simple task.

‘

This is one that is both practice and actual design, and I’ve not yet decided how best to fit it to the shaft. So the hole through the middle is nominal.

The one photo I have of it, shows a large central nut so the wheel was probably on a taper, serrations or keyed. I doubt I can cut serrations accurately enough for them all to register correctly, and anyway the shaft won’t be very thick. The present steering-gear is temporary, using a second-hand, rather small worm and wheel. The steering-wheel is too small. It is held to the existing shaft by a nut on a threaded extension, with a roll-pin not across the diameter but down the joint, as a key.

Parts like this are rather prominent and I don’t know their original full-size dimensions, so have make educated (?) guesses. I need determine what I can from a few ancient photos to derive the model size, then multiply that size back by three to consider how realistic it is. Rivet-counters would have kittens!

So in this example a half-inch diameter spindle, which will run in a hollow column, seems very slender but gives a full-size inch and half diameter, which seems excessive. Similarly, once I’d thought the miniature steering-wheel’s diameter ought be 4.75″, I held a tape-measure in both hands open at the 3X length (14.25″ ) to see if that felt about right.

[Author]

Posts