Member postings for Nigel Graham 2

JasonB

According to the e-posted notification you invited me to send an example, but I can't see the invitation here.

Never mind, thank you for the invitation.

I don't know if I can take you up on that, I'm afraid, but the machine I am trying to design is the enclosed, inverted-vertical compound engine for a 4"-scale Hindley steam-wagon.

It took me a long time to find a way to design it, irrespective of how to draw it, with no original drawings known. I have now found a proven miniature engine design as a template for the motion-work's geometry; and have started to draw the assembly and parts.

So far, I have drawn the crankshaft from scratch. I'd already part-machined the cylinders so have drawn them from their measurements, and presently I am designing and drawing the engine case.

Geometrically, so far this is a fairly easy set of rectangles and circles, and all in 2D mode.

I had drawn the crankshaft and even started trying to draw the engine's GA, both in 3D. I found both were long and laborious tasks; but the latter soon proved impossible.

I found out how to place that first version of the shaft on the TurboCAD forum gallery as a first 3D attempt, in outline only. Another user kindly coloured it for me! It looks pretty but that's about all.

As a last straw, after my becoming disheartened by a beautiful Alibre rendering of a marine engine on a recent ME or MEW cover, the TurboCAD User's Forum log-on no longer works, at least from my computer. So far IMSI has been unable to repair it, and it prevents my seeking help.

To SillyOld Duffer.

I'm afraid you are reading too much into my motives and needs.

I am trying to design things I can make with my workshop and abilities. I am not in the plans business, nor making brochures. I don't really believe in phrases like "world view", either.

I can see the point of 3D CAD images, and I know CAD anyway requires a new approach of its own, beyond manual draughting.

Vicarious experience at work showed me the advantages of CAD, and I bought TurboCAD. It was the only genuine CAD package available to the amateur at the time - only a few years ago .

I sampled Fusion360 (like Alibre, 3D-model-then-draw) - gritting my teeth at its "in your face" presentation - and AutoCAD 2000 (2D-only I think). SolidWorks is not intended for amateurs. Of various low-budget or "free" CAD packages on-line, those not devoid of engineering-drawing facilities anyway were heavily-stripped "student" or "trial" editions of costly industrial software.

After a monumental struggle I can make good-enough-for-me orthographic drawings, and very, very simple 3D images; but much further is beyond me, including accurately placing and linking multiple 3D elements. (For simple cases and symmetrical elements, I calculate their co-ordinates, but that is slow, error-prone, poor practice.) I will have to live without 3D CAD despite having hoped originally to be able to use that facility.

So I tried Alibre, hoping it might be easier to learn than TurboCAD despite doing the same things. I think I missed an edition of MEW so that won't help, and so far only have the scribing-block base from Part 1 of the series. Given all this, it is not logical to start completely anew on a very different make of CAD with a 3D-first approach.

Thankyou Jason.

Yes, I did read the verb "might", and I did try other file types including ones with the magic letters 'DW', but without success. Given the very long list of possibilities though, it would be very easy to miss the one that might work..

Neill - I did spot your accidental use of "project" but wasn't quite sure what it ought have been.

I'm familiar with the "extrude" principle, as it's a basic one in TurboCAD 3D too. TC though gives several different ways of producing a "solid" - extruding from a planar figure, and editing a "primitive" ready-made object being perhaps the main ones; and I spent lot of time experimenting with them.

I never reached the advanced levels of "sweeps" and the like, but I found even with those two basic actions the properties of the generated solids and their reactions to editing tools differ markedly, with no clear rhyme, reason, pattern or manual instructions to help you choose the right type where both will give the same figure (mainly cuboids and cylinders). I assumed this will be common to all 3D CAD packages, because they have to perform essentially the same calculations and routines behind the presentation.

The MEW example works because all such decision-making has been done for you, so if you follow the instructions carefully you end up with a prettily-coloured picture of scribing-block already designed, without you really understanding how. It's a rote technique, valuable as a primer, but if Alibre has the same subtleties as TurboCAD and no doubt "others available", a rote primer won't guarantee future success with your own designs.

I still have the drawing of the scribing-block's base from Part One - I even succeeded in enlarging its diameter and recessing its underside so a real version would be more stable.

[That's as far as I reached. I had bought the magazine in Smiths, and took out a subscription (itself not as simple as it had seemed) to start with the next issue, but I think it jumped one. I have not yet verified edition numbers but the Alibre series appeared to refer to things not in Part One.]

I do realise advanced CAD skills can let you make images that would be horrendously difficult to draw manually; but that can be a trap, as the components would become horrendously difficult to make on an ordinary lathe and milling-machine with normal model-engineering skills.

Don't forget our own projects as model-engineers are either new to our own design and can use modern design practices, or are replicas of machines built in the days of exquisite draughtsmanship by pencil, T-square and set-square. Either way they have to suit our own workshops and craft abilities.

Having said that, I do know a lot of model-engineers now are using small CNC milling-machines and 3D-printers, so provided they still ensure the design suits the machine, the problem of physical difficulty is reduced somewhat.

(I recall seeing my first CNC mill in the factory where I was a storekeeper. One of the work-pieces for which it was intended was carved out of a big aluminium-alloy slab. The first attempt revealed it too large for the machine, which itself was sizeable; so while that batch was milled conventionally the drawing-office re-designed the part to fit the CNC mill! Pencil and paper designs, too.)

You still need to understand both Design and CAD. They are separate. The latter is extremely difficult, and exists to express the former as images.

Errr, thank you SiilyOldDuffer but with great respect I can't quite understand what you mean, beyond your having totally misrepresented me!

I am not blaming the software either.

I have no idea what you mean by "world view" but I tried all 3 CAD programmes - plus AutoCAD 2000 - to see what I would find the best for me.

Neill - Sorry, but your comments about "parametric", "object trees" etc are over my head. I think they are for advanced users.

Your second paragraph does not apply to me. I am used to orthographic drawings by default and am not trying to "project" them anywhere; but I use isometric as an auxiliary tool.

David - I tried to import a TurboCAD file into Alibre, to see if it's possible. I found Alibre does not read TCW files.

I appreciate that CAD generates "solids" from planar figures, but in different ways giving the same shapes with very different internal characters and reactions.

It took me a long time to realise why I could not draw a simple cross-section. Apart from the term not being used anyway, the programme did NOT tell me why I could not make it work.

Those last two observations refer to TurboCAD, but I am not convinced any CAD programme would be any better.

I understand the notion of drawing a model by adding geometrical objects to each other, but 3D uses many ways to do so, and it is by no means obvious which to use and which not, how and why. Anyway, an orthographic drawings is also a set of objects stuck together. The only difference is in using a single work-plane.

Alibre, TurboCAD and Fusion360 essentially do the same things though in different ways. Fusion initially looked simpler than TC but I found it's about the same level of difficulty. It also seems to force the 3D-then-2D route as Alibre does.

TC offers you the 2D/3D choice from the start, if you do not need to draw in 3D. Orthographic-only is still difficult but at least avoids its 3D's incomprehensible morass of combinations and permutations of multiple "solid" generators, work-planes and co-ordinate systems. I don't know, but suspect Alibre and F360 will have comparable systems, as they all do fundamentally the same things.

Following a series to make a specific drawing of some specific, fairly simple example teaches you how to draw that example; but whilst useful you need far more than that to use CAD properly and usefully even for similarly simple items.

A video only shows you an expert demonstrating the software.

My requirements are simple:

1) A CAD programme that helps me to design and draw particular engineering items so I can make them.

2) Isometric: perhaps useful for clarifying awkward shapes and assemblies, but not essential.

3) The programme being reasonably easy and intuitive; with proper operating information available (some hope).

Rendered pictures merely add add a huge extra burden of skills, knowledge and work to the process.

Being able to make a prettily-coloured model is not itself a reason for doing so.. By corollary, recognising the limits of own abilities are a reason for not attempting that. I cannot learn how to make 3D CAD pictures, but I do not need to anyway, and do not see the why Alibre wants me to use a roundabout way to make the real drawings. I want to make metal things, not "paintings".

I have explained why I cannot change to Alibre now. I am perfectly well aware these programmes all differ significantly in how you use them, and there is no point in abandoning one I know enough to create admittedly-rough drawings, and start all over again with another!

Further, I now have quite a number of TurboCAD drawings, but Alibre does not recognise their file-types so cannot open them. I investigated this only today.

It seems I've found a gulf of understanding besides that already between CAD and me, but also between my CAD needs and others' preferences in using it.

Thank you David.

15-30 days would be fine for a CAD expert to evaluate and review the programme. I would say it needs at least 6 months of very regular use to learn even fairly simple CAD from scratch.

I find what is lacking is any decent printed material. If you come to CAD cold from manual drawing, you are are faced with a huge knowledge gap. You understand making and using engineering items and their conventional plans, but CAD places an enormous extra skill gap between designing and making.

There is virtually nothing available to help you bridge that gap.

The software publishers assume formal, professional training in industry, so do not write manuals beyond ones like TurboCAD's pdf aide-memoire. They might produce training videos, but not everyone can learn from them. Further, the dedicated "manuals" assume you already understand CAD concepts as well as the engineering and draughting knowledge, and in depth.

Most big software firms duck responsibility by establishing "users' forums" and "[In]Frequently Asked Questions". Well, why not write decent instructions in the first place? Is there an Alibre book for example? (There isn't for TurboCAD.)

You can't just write "Select object X, click on Tool B". You need to know why Tool B, and crucially, what existing but hidden conditions in X are needed for B to work, such as mathematically rather than image-sight accuracy in joining and trimming, perhaps created much earlier,

I have two books that bravely try to convey the basic concepts. These, by D.A.G. Brown and by Neill Hughes, help, though I found the latter's "meter" for "metre" inexcusable and annoying. They explain some general CAD basics but still do not warn of the pitfalls. Like the CAD publishers, they assume you never make mistakes, and if it's in the computer it must be right. In fact CAD lets you unwittingly make those hidden mistakes I cited, because they do not stop the application itself; making the whole thing potentially an extremely frustrating waste.

So why the paucity of proper help in CAD? It's partly a commercial matter, but years of work and hobby experience has taught me that the world is full of experts in many fields, especially but not only IT, who understand it easily and almost instinctively, and cannot comprehend why mere mortals find it hard to learn. Instead they show what they can do and tell you it's easy (one reason I don't believe in "training videos", and displays of highly-detailed CAD engineering and architectural artwork discourage rather than encourage.

So why do I bother to use/waste hours of potential workshop time learning CAD?

I thought it might help my model-engineering!

Most model-engineers work from published plans so might never need to learn CAD except for its own interest or if they wish to send CAM files to materials suppliers - or indeed use CNC in their own workshops.

However, I am trying to build a large-scale miniature steam-wagon from no more than scraps of archive material, no castings, much "recycled" material. I have to design as I build - lots of traps and reversals. So I could see the advantages of CAD's drawing, editing and copying abilities.

I could not have known CAD is so difficult and frustrating. I had to abandon the 3D nightmare, and stick to relatively simple orthographic projections. Anyway making parts needs 2D elevations, not pictures. For model-engineering, isometrics are useful for assembly-drawings but not essential, and rendered pictures really show only an undoubtedly high but extra skill in making rendered pictures.

Would I use Alibre? Probably not now.

If starting from scratch, probably yes. It would have been sensible thanks to MEW's trial version and course. I have installed Alibre and I followed the first episode.

However, I've probably gone too far with TurboCAD to change now. It is a completely different method for the same task; I do not know if Alibre can take my existing TC drawings; and the Alibre + MEW, 3D-then-2D way, is wrong for me.

Thank you David.

The e-posts Alibre sends don't bother me too much, though their style is sometimes grating. I just delete them.

I realised very soon that Alibre would be no easier to learn than TurboCAD, even if I was mistaken thinking some tools had been switched off for the MEW course.

My problem was that Alibre was presented as a "model then draw" system. I recall a TurboCAD forum contributor stating TC is meant to be used in the same way, but I cannot learn its 3D side. I'd found similar with Fusion360. So I assume that all modern CAD programmes are much the same in intent if not methods.

3D CAD is extremely complex, and lacks useful manuals; but I want to make workshop drawings not brochure "paintings". However,I took up CAD partly for its facility to produce rotatable isometric assembly-drawings. Please don't get me wrong though: I respect the skill others master even for something as fairly simple as, say, the parting-tool holder that keeps flashing up in the side-bar ads. I can't play the piano, but I can still appreciate a Romantic-era concerto.

I made a mistake. I tried Alibre despite having already gone somewhere vague with TurboCAD: quite different but equally difficult. Worse, the MEW introductory course is straight into 3D CAD, which I'd already found too hard.

So on balance, I think it best to decide first if CAD is likely to be both feasible and helpful for you, and if so, to choose and learn just one make of CAD, to your own limit. If that limit is very modest, so be it, just accept and keep within it.

I'd wondered where Copy In Place had gone, but found it disguised as a pair of footprints as a symbol.

I use it a lot, then moving or resizing the copy as needed (such as concentric circles).

I take your point about security, and I am sure Fusion is as careful as the banks. Nevertheless, I do not trust the overall security of the Internet as its attackers now are very highly skilled IT professionals paid to keep ahead of any security precautions. I would not worry, if I ever try Fusion again, if my rough CAD attempts are on the www somewhere, but I am not running a company in sensitive work.

That company concerned was in defence manufacturing, so could not take any risks.

It would make neither commercial nor security sense for any company handling very sensitive information to put the control of any of its administrative and intellectual property outside, beyond its own building and firewall. The way things are going with the Internet, we might not need worry about Foreign Powers snooping on our model-engineering drawings, but in the business world, I foresee future consultants not trying to persuade firms to make their IT people redundant to save a few hundred quid, but advising just the opposite.

Especially when the firm's Intranet handles things like technical reports and design drawings too.

Besides, it may be cheaper but that doesn't necessarily mean "better" anyway.

Alibre in MEW:

It seems to be a simplified "trial" version. I worked through the first instalment, then as invited tried exploring it further, bringing a very sparse and patchy knowledge of CAD via still fighting TurboCAD and being defeated by Fusion360.

I soon found it not possible to place objects where I wanted, and close examination of the printed screen-shots and the tool-bars on screen, revealed some significant commands had been switched off. (Faded symbols)

I thought, "Oh aye? I wonder what it will cost when the 6-month trial ends if I want to use it for real and need such details as specifying co-ordinates?"

So by all means try it and buy the full version if you find it suits you, but be aware the MEW tutorial is using a stripped version.

If you are yet to try it, Alibre's publishers also regularly send you irritating, somewhat patronising and frankly pointless "marketing" e-posts!

I did find the Alibre course moderately easy as far as I went, but I was following the recipe in MEW, and in due course decided having gone so far with TurboCAD I was being daft to start again.

To be fair to both programmes though, I ought point out though my approach was totally different; and I know my own limits with CAD. The Alibre / MEW course says draw a 3D model then produce the orthographic workshop drawings from that. In TurboCAD, I stick to orthographic-only as its hellishly difficult 3D modelling is beyond me. Let's be honest though; software like TC, Alibre, Fusion and SW is really intended for professional users given costly, professional training. If you don't have the aptitude for teaching yourself industrial-grade software for which no proper manuals exist, you won't learn it to any very useful level.

I don't think I will be disposing of my industrial-grade, A0-size, parallel-motion drawing-board!

Files:

I know I've come into this a long time after the last post, but I am fairly certain that Fusion does allow you to store your drawings in a folder of your own if you wish. Its use of the Internet - called the "cloud" I suspect to camouflage it slightly - is its default.

Admittedly, it's unlikely to worry us too much, but I once asked a design & manufacturing company's IT manager if a firm like his would ever consider buying any CAD programme that uses the cloud for data files.

He laughed, and said, "Not likely!".

Anyone here use TurboCAD, which was being sold through the model-engineering trade but apparently no longer.

I can open its Users' Forum but not log in, which is a bind because I'm stuck on making the dimensions work properly. Just in orthographic - not 3D.

It was always flakey, often wanting me to repeat the password at least once, and every few random sessions it would forget completely and accuse me of doing that. It now fails to send the re-set e-mail it claims it has.

I found, eventually, a link on the main IMSI site (software manufacturers don't like you to ask anything except "Please May I Buy...?". They sent a password reset thing but it achieved nothing and I think only refers to the IMSI site. I had the impression they didn't understand what I was asking.

I have altered nothing on my computer, so has anyone else found the TC Forum now inaccessible with no means to regain it?

I have an ancient Denbigh Horizontal Milling Machine, and would love to know more about it! It far pre-dates anything about Denbigh on Tony Griffith's 'Lathes' web-site.

Table size about 17" x 5" -ish. Vertical travel about 7"? (From memory - it's presently stored in bits, awaiting rebuilding so I can use it!)

3MT spindle in plain bearings.

Unusual lead-screw pitches: 6tpi on 2 of them, 8TPI on the other.

Power-feed wormwheel fitted but no drive-shaft - either lost or an option not taken by the original owners.

Built for overhead drive to 3-step flat-belt pulley - the casting forms a loop so you can't use a V-belt upwards.

Single, round-section, steel bar over-arm. The broken drop-bracket that came with it looks like some substitute from a totally different machine, not an original at all.

Bench-mounting but has its own, very heavy, one-piece cast-iron stand with integral, shallow chip-tray and large open central cavity so it vaguely resembles a Victorian loo!

No obvious model type or number, apart from the name, Staffordshire Knot trade-mark and the digit "4" stamped into some machined surfaces - possibly a batch number rather than serial number.

I have seen what I think is a variant on the same mill, with lever-action traverse the text said, in a photo on ME not long ago, but it was largely hidden by its owner's set-up to use it as a faceplate lathe for machining miniature traction-engine parts.

Small, early horizontal-mills were quite often fitted with lever feed for mass-production manufacturing by semi-skilled operators. The 6TPI threads on mine may have been to special order - I just don't know!

Headstock Bearings: Stick with replacing like-for-like. Apart from being much simpler to do, fitting races may need the casting boring out so much it will be seriously weakened. Also of course, if you sell the machine on in future, it may be more valuable for being original.

If you have the mandrel re-ground you'll need to make up new bearings to suit. Are the bearings adjustable? If so you may able to tighten them enough temporarily to make the new liners, having first measured the new diameter & made a Go/No-go gauge.

A better alternative would be to make the new bearings on another, good lathe of course!

Change-wheels: I asked Myford this very question as I have a 1908-pattern Drummond flat-bed lathe in need of TLC. Drummond & Myford wheels will not mesh properly. They may have the same DP & bore but apart from different keying, the pressure-angle is different. Try to find the original wheels' specs and match that; or make up a complete set separately.

\Warning: Drummond change-wheel pins are in tapered holes and driving the pins the wrong way can break the wheel.