The Home Machinist!

As I am taking his original dimensions and plans and reconstructing his Petrolea engine in 3d CAD, a few things have become extremely apparent. For one thing, I do not believe he ever actually built the Petrolea. This is very evident in the numerous gross errors in his drawings. I was warned that his drawings need to be looked over with a fine tooth comb, and it has proven to be true.
For one example, the crank pins on the drivers for the coupling rods actually strike the outside frame, preventing the drivers from being able to rotate.
Another example is that the rear drivers cut into the rear buffer beams by a considerable margin. The dimensions of the wheels themselves do not even closely relate to the Wheel Standards we use today(I can't fault him for that).

I was told of one gross error that I have not stumbled upon yet, is that the dimensions of the boiler for the Petrolea will not fit between the frames of the said locomotive... Imagine if you built this then had the boiler made elsewhere without checking dimensions?

No matter where the drawings come from, you best check all dimensions before you make any big purchases or cut any metal!

welcome to the world of live steam models.

It wasn't all that many years ago that a friend and I finished a loco, but had to totall redesign the valve gear since there was no way, as designed, it would actually run.

This from one of the more 'respected' suppliers of the time.

There is a photo of "Petrolea"on page 69 of LBSC's biography: "LBSC" his life and locomotives by Brian Hollingsworth.
It says there that the series was the second longest running serial published in the Model Engineer. No pictures in the serial?
In the book it says that LBSC build locomotives without drawings, so it was up to others and at the time of the book Kennion appears the sole Petrolea drawing supplier.
Please note that Alan Stepney, also on this forum, maintains an "error in drawings list" please supply your known errors to him.
If somebody else had done that earlier it might have saved you time!
Kind regards
Jos Koopmans

Even with all the errors, I still find it amazing that he could design all those engines without building them (designed about 100, built about 50, so there are about 50 more that were all on paper).

All this with pencil and paper. Most without a calculator. All without CAD.

Steve

With some of his locos, Curly built one then either did the drawings or gave it to someone else to draw.

Juliet was done that way. Donaldson actually drew it, based upon an engine LBSC had built.
(And a fine engine they are too.)

As Bill said, errors are not unusual with model drawings.

Check, check and check again, and you may well still have some mistakes!

Thank you for the tip about the error log, I will be sure to pass on my information.
I have respect for LBSC for being such a prolific designer back in the day, which many of us today are still building his locomotives.

LBSC wasn't alone in this. Martin Evans, for many years the editor of Model Engineer, did not build a large percentage of his loco designs. He typically sketched new designs and handed them off to the ME drawing office (at one time a couple of very talented ladies) who turned them into publishable drawings. Shortly before I ended my subscription to ME Evans began a new locomotive series which I was very interested in, an LSWR T6 4-4-0, but he mentioned at the outset the series would be short, and there only be the most basic of drawings, enough to give the builder a push in the right direction, and the details could be filled in from previous designs.

Evans was no better or worse than any other designer in that his drawings contained errors and there was hardly ever a series where an occasional construction article wasn't followed in a few weeks by a correction. Aside from this the one criticism of Evans I've read complained that many of his designs contained recycled details which often weren't suitable for the particular design, or didn't follow the appearance of the prototype, or both.

As has been mentioned I think Alan is providing a great potential resource for the model engineering community, with his Errors website page, but it can only become a great resource if folks take the time to report errors.

Even Kozo Hiraoka has some mistakes that have crept into his books, there is an errors thread on the DLS Kozo forums where we keep track of them all. Kozo built all of his engines except the New Shay, for that one he sent drawings to a trusted friend as he finished them and the friend built the first New Shay. Kozo no longer has a machine shop but he is still drawing. As to what he is drawing I am not allowed to say just yet.

We found a number of problems with Gene's Chloe drawings, or missing critical dimensions in some cases. If you expect them you can work around them, but you really do have to pay attention.

In making up the formal drawings for the Darjeeling B-Class, I am having multiple people review them, and they are also based on an as-completed model, so a lot of the bugs are being worked out beforehand. Even then the subsequent first few builders (including myself) using the drawings will find things, and hopefully just minor issues at that point. Modeling in 3D helps find interferences but it's still easy to make mistakes. having formal drawing review sessions really helps.

Another anachronism that a lot of the older drawings have is fractions. I personally don't like them and all of my dimensions are decimal. Anyone have a particular opinion about that?

Harlock wrote:Even Kozo Hiraoka has some mistakes that have crept into his books, there is an errors thread on the DLS Kozo forums where we keep track of them all. Kozo built all of his engines except the New Shay, for that one he sent drawings to a trusted friend as he finished them and the friend built the first New Shay. Kozo no longer has a machine shop but he is still drawing. As to what he is drawing I am not allowed to say just yet.

We found a number of problems with Gene's Chloe drawings, or missing critical dimensions in some cases. If you expect them you can work around them, but you really do have to pay attention.

In making up the formal drawings for the Darjeeling B-Class, I am having multiple people review them, and they are also based on an as-completed model, so a lot of the bugs are being worked out beforehand. Even then the subsequent first few builders (including myself) using the drawings will find things, and hopefully just minor issues at that point. Modeling in 3D helps find interferences but it's still easy to make mistakes. having formal drawing review sessions really helps.

Another anachronism that a lot of the older drawings have is fractions. I personally don't like them and all of my dimensions are decimal. Anyone have a particular opinion about that?

I also do not like fractions. All the drawings when implanted into solidworks get converted to decimal. I think by using Solidworks first to build a 3d model, you find all the critical errors. The secondary errors come up when you make the 2d drawings and leave out certain dimensions or have the wrong tolerances to certain parts requiring slip fits or press fits. Granted, LBSC only used fractions on the drawings I have seen, and absolutely no tolerances listed on any drawings, just what he hides in his writings. Further proof that he actually did not have anything to do with the drawings.

RB211 wrote: I also do not like fractions. All the drawings when implanted into solidworks get converted to decimal. I think by using Solidworks first to build a 3d model, you find all the critical errors. The secondary errors come up when you make the 2d drawings and leave out certain dimensions or have the wrong tolerances to certain parts requiring slip fits or press fits. Granted, LBSC only used fractions on the drawings I have seen, and absolutely no tolerances listed on any drawings, just what he hides in his writings. Further proof that he actually did not have anything to do with the drawings.

I'm doing the same process with SW, so right there with you. SW takes out a certain fudge factor, and it's down to the person using it to make the parts correctly and then make good drawings in the drawing editor. I'm trying to follow recent ANSI standards gleaned from books. SW's default annotation settings are not always what is in the most common use, so I've had to tweak them a little here and there.

An example is when you insert a projected view, the view can face away from the parent view or towards it, and there's a definite preference for one or the other depending on who you talk to. SW does the opposite of the standard we use at work by default.

As has been mentioned, LBSC built his locos 'out of his head' and did not make any drawings at the time. I'm pretty sure that he did produce the drawings for all of his designs after they were built. Unfortunately, these were often done from memory rather than measurements from the finished product which explains why some of the errors crept in. These original drawings were then often reproduced by other bodies such as Donaldson, Kennions etc. and further errors were introduced in the translation. When the construction articles appeared in Model Engineer and English Mechanics, LBSC's original drawings were sent to the magazines where they were copied by the illustrators etc. and a few more errors were probably introduced. At home I have a set of LBSC's original pencil drawings for Rose, a single cylinder 2 1/2" gauge loco which appeared in English Mechanics. They are works of art despite his insistance that he was no draughtsman.

It is extremely difficult to produce a set of 2D drawings that are error free and virtually impossible if you don't actually build a loco etc. to check the drawings you have produced. 3D does help but is still not 100% without proving the design by building it.

John

Harlock wrote:Anyone have a particular opinion about that?

Likely not one of concern.

Fractions were a good way to display dimensions that weren't critical, thus they had a wider tolerance. Two place decimal dim's = ± .010", three place decimals = ± .005" and fractions ± 1/64", or ± 1/32".

They still have their place, and serve to keep you sharp in remembering conversions to decimals.

Harold