Lempor Draughting (or Bodgers Blastpipe) by Bill Roebuck

I had intended to write this article on improving the steaming of an engine some years ago, but retirement and house moving got in the way, and the text quietly disappeared into the mists of 'things to do one day'. The investigations into better draughting were prompted by the completion of an LBSC 3%" Netta (photograph 1), which did not run or steam as freely as it should, and the blast sounded 'tearing'. Something would have to be done. As an aside, it may be worth checking your is completely airtight so your existing arrangements have the chance to work correctly before embarking on this Photograph 7: The Author's LBSC Netta as completed. journey. A tiny hole too small to stick a pin into can completely destroy the smokebox on for boiler feed reliable steam My research implied that the way to vacuum, and no amount of blastpipe was essential. go was with multiple jet blastpipes of the tinkering will make much difference. Guess Much reading of books and magazines LemPor variety (Lemaitre meets Porta). The how Ifound that out! followed to try and get a better idea lemPor blastpipe comprises four De Laval Some years ago an article in Model of which way to go in improving the style nozzles set in a cross, all leaning slightly Engineer suggested that steaming could draughting. The main consensus seemed outwards (photographs 2 & 3). The four be improved by making the blast nozzle to be that a properly designed single steam jets would certainly fill the 15.5% of the cylinder diameter, then blast-pipe/sinqle-chimney arranqement choke, although LemPor draughting is drawing up the blast cone to an induded should be capable of making all the steam usually used with a draught tube. This is like angle of 60 and positioning the choke 50 required, but the ideas, dimensions and an extended petticoat pipe (photograph 4), the generated cone filled it (or adjusting proportions required to achieve this were but Ithought that improvements might be the orifice: choke height ratio to between often vague and mostly contradictory. One achieved without this. Some doodling on 1:9 and 1:10). I suspect these proportions designer made frequent reference to the the drawing board soon produced what dated back to Henry Greenly, but I have to American Master-Mechanics smokebox looked like an acceptable set of nozzle admit that the re-draughting of a previous proportions, without ever actually saying proportions based on a bit of rough scaling LBSC in this way resulted in what they were or how to calculate them. I from photographs of full size ones being the engine becoming a consistent and did extrapolate from the dimensions of one machined (photograph 5). Some form of predictable steamer which ran much more design reputed to steam well and made jig would have to be made to machine freely due to the lower back-pressure. the required bits to suit Netta, but I think the nozzle bases so they would sit in a Netta, unfortunately, did not respond to it might be best to draw a veil over the cluster for brazing together, and tapered this treatment. You couldn't say it didn't result. I still get asked whether I subscribe reamers would also have to be made, but steam, but it certainly didn't steam well or to the 'get-you-home' service for steamless apart from that the process looked fairly reliably, and being at the time dependent engines. straightforward.

Photograph 2: A set of Lempor nozzles at the Photograph 4: Fabricated draught tube for W & Welshpao( & llaniair Rly. L Beyer Peacock locomotive. Photograph 3:As photograph 2 but with a blower fitted.

ENGINEERING in MINIATURE December 20 f 1 o I\) i I 3D '1 0' ..,I\) N STEP 1 SCRAP VIEW C) ••• DRILL NOZZLE SHOWING ELEVATION ••• ~'---':--ES£~1~AND HALF SECTION OF BLAST CAP

( ~JZZZZZZZZ?(4-+' STEP 2 RELIEVE MINOR REAM 4° TAPER \ DIAMETER LOCALLY PART OFF T IF NEEDED SILVER SOLDER 2/2D PARTOFF~ \ ~ C\- ~~;~~~:[L~f~~4~~~s~%P:~i':pf-::72:~;;t;p·r I.. D 1 2D ,I ; THREAD TO SUIT -. = BLAST PIPE !: STEP 4 NOZZLE CLUSTER Si REAM 8° TAPER TO LEAVE : .. ~ MAJOR DIAMETER 5; SHORT FLAT MAJOR DIAMETER RECESS. NOZZLE CLUSTER {~:=-:;;~;11f-T MINOR DIAMETER MINOR DIAMETER ~ RELIEVED AS SHOWN. 5° I 1%D I (NOZZLE BASES

I I OUTLINE SHOWN) STEP 5 CLAMP IN JIG - MILL ONE FLAT AT 5° APPROX \ L------J- --~--

I 1

1 I STEP 6 ROTATE 90°, PRODUCTION SEQUENCE FOR CLAMP IN JIG· MILL SECOND NOZZLES TO GIVE LEMPOR FLAT AT 5° APPROX. i2-2:------} CLEAN UP. --$--

1 DRAUGHTING Left - Photograph 5: Lempor nozzles in process of being machined .. Note taper is (anger than on the miniature set.

Right - Photograph 6: Side view of the model milling jig.

Purists and 'proper' model engineers calculated as above becomes the outside they provide a table I will look forward to will no doubt wince at the imprecise nature diameter of each nozzle. Well, sort of. I using their result. but unless you make your of what follows, but SMEE does stand went for the next 'standard' size above, own nozzle choke drills you are still tied to for the Society of Model & Experimental and had no problems. Whatever you number or metric equivalent twist drills. Engineers. If my ad-hoc unscientific decide, the nozzle choke diameter is half approach undoubtedly seems a bit rough the outside diameter, so the total area of Task 2: Make the Reamers and ready, then that's probably because the four nozzles is the same asthe area of a The D-bit style reamer angles of 4 and it is. Mathematics, geometry and I met single hole. As an example, if the calculated 8 degrees were a compromise to give a 'fit' at school, recognised we would never blastpipe size is 1,4" (D),nozzle OD is 1,4" (D), from the drawing. They were machined understand each other, and passed on like choke dia. W' (Y2D). Nozzle length should from silver steel in the usual way by setting ships in the night. If my teacher had been be about 3D - i.e. 3 times the nozzle OD. the top slide round. There does not seem granted the foresight to teJl me I would This works for nozzles up to about 3/S" dia. any criticality in the angle - use the degree need to know about these things so I Over that you are a bit on your own but graduations provided. The critical factor is could build steam engines in my old age it should be possible to extrapolate for in the depth they are used - it is important I might have paid more attention. As it is, larger values. I have to say there does not to leave the short parallel portion in the guesswork and empiricism suffice. If this seem to be any magic to the design size nozzle. You could include this parallel part offends you I apologise, but I am sure the that you end up with. Exhaustive testing on the end of the taper portion to save editor would be pleased to receive your has shown that the odd 64th up or down having to set depth stops, but it will mean mathematical treatise on the design of makes no difference to the end result - the you need a separate reamer for each size exhaust systems, then we could all get it spread of the nozzles ensures the chimney of nozzle. See the paragraph regarding right first time. choke is filled, and the smokebox vacuum the machining jig before altering the top should automatically follow. No doubt the slide back to turn parallel - the nozzle Task 1:Select your nozzle size mathematicians out there will pick me up taper wedge clamp needs to be machined This is straightforward (at least up to on this, with complicated trigonometrical at the same taper angle as the 40 reamer. about 0.375").The equivalent single orifice formulae to calculate every dimension In deference to the importance of the job blastpipe size either from the drawings or down to three decimal places. Providing I machined the half diameter down on the 7BA J';f- T~\¥ ..~ ..- ~T- ,.7j( I

------TAPER WEDGE

NOZZLE MILLING JI.G %" SBA I. THIS FACE 112" MUST BE 90°

~ S;:t ~'I {@tr ~-1 ..- ~rr--+-----J ~ c~"%" ~ I· %••

CLAMP PLATE 1"

ENGINEERING in MINIATURE December 2011 Photograph 9: Turnednozzle blank clamped in the milling jig (step5).

Photograph 7: Top view of the milling jig. Photograph 8: Turningthe 4" external taper (step 2). Perform the above 4 times for each set of nozzles to be made. I found it easier to mill, but [ think this was to give my ego the larger nozzles you could increase this - Le. turn and ream all the nozzle blanks at the impression I was working accurately - you SOA if your minimum nozzle ID was going same time for the three sizes I had decided will no doubt file yours carefully down. to be 1/a". Brasswould be a good choice, on (calculated size plus one size above and Harden and temper to straw (any shade but steel will give you a better thread for below). To save grief later, keep each set of - it's only going to be used on brass) - clamping purposes. As they are to some nozzles of the same size together in a little probably best to do all the quenching by extent sacrificial if you are a cautious-type, bag, as teWng them' apart once finished is plunging vertically into the water or oil to or are anticipating mass production of frustrating. Photograph 5 shows some full minimise warping. nozzles, maketwo or three whilst you are at size nozzles at various stages of machining. it (photographs 6, 7 & 8). You will also need Yoursshould look a bit like these. Task 3: Make the Machining Jig to make a pull rod for the taper wedg.e Step 3: Takeeach parallel blank in turn, The only thing to decidebefore making (not detailed). Mine is 3/32" steel about 1" grip in the chuck by the reamed end, and use this is the maximum size of nozzle you long with some 7BA thread on each end. the top slide to turn the outside taper with anticipate making. The dimensions shown As usual it is too long for the smallest size light cuts and a sharp tool (photograph 9). will cope with up to %" nozzle 00, which is nozzle (hence the thick washer) and just If you are really picky you will measure the ample for 3W' gauge. If you want to make about long enough for the largest size. taper as you go, but I did all mine by eye. big nozzles for 7W' gauge engines you will The main thing isto leave enough 'meat' for have to increase the dimensions marked Task 4: Make the Nozzles the 8° reamer. with an asterisk to suit. To 'prove' the Making nozzles is very easy, if a little Step 4: the depth of the 80 internal design I'made a set of ~/8"dia nozzles, and tedious. First set the top slide round to reverse taper is a bit important, as you machined the flats on a temporary fixture. about 4°. If your 3-jaw chuck is reasonably should leave a short flat between the tapers. It did work, but I would not like to have to accurate then use that, otherwise use the I did the first one by eye, then used the do it again. The nozzles must be securely 4-jaw - the material needs to run about tailstock graduations to ensure consistency. clamped for milling the flats. centred or you may ream your way out of Take this step gently - your hold on the The jig as drawn can be cut from an 1" the side of it. nozzle blank is precarious. If you would feel square off-cut of steel, but any size near Step 1: Chuck your selected size of happier set a reamer depth gauge first. this will do. It needs to be rigid enough brass rod with about 1" projecting and if Step 5: With all four nozzles turned not to flex when clamped in the mill vice. using a 4-jaw adjust to run true and from set the machining jig up on the milling Do not be tempted to machine a jig to now on only slacken and tiqhten the same machine (or whatever method you use for compound angles so as to machine both adjacent 2 jaws - I always use 1 and 4. Yes, milling). Clamp the first nozzle in as shown flats at one setting unless you like throwing I know this does not automatically ensure in photograph 10, and using a new, sharp away spoilt nozzles. The drawIngs show the re-chucking exactly concentric, but it will end mill gently mill the first flat. Maximum other bits needed - the end of the clamp be near enough for our purposes. Well cuts here of 5 thou - less if you are of a iron must be at 90° to the face. I made my equipped engineers will use a good 3-jaw nervous disposition. Being heavy handed at clamp iron W' thick, but this is not really or collett. Centrethe blank, and drill though this stage means lots of naughty words and thick enough so I have shown 3/16" on the with your choke size drill. a spoilt nozzle - you have been warned. It is drawing. The slot should be a fairly good Step 2: Estimate the depth of the main difficult to give exact figures for the amount fit to the clamp screw. Turn the nozzle taper (about 2'120), set a stop on the reamer to be taken off for each flat. Aim to get a taper clamp to the same nominal 40 taper if you like and then gently ream the nozzle crescent-shaped cut-out showing like the as the reamer (see paragraph on making to depth. Aim for a good finish but don't illustration (photographs 11, 12, 13), with reamers), and tap 7BA. If you are making panic if it's not. Part off to length. the flat about 3/5ths of the length of the nozzle. Repeatfor all 4 nozzles in a set. Below - Photograph 10:After milling the first flat on a nozzle. Step 6: Slacken the taper clamp, turn the nozzle through 90°, take up the taper clamp enough to hold while you tighten the clamp iron up close to the milled flat. Then tighten the taper clamp, and mill the second flat to the same setting as the first. The two flats should cut into the nozzle to leavethe 'point' as shown just above the nozzle base in photographs 14 to 17.Within reason there is no major criticality here - a few thou either side of optimum will be perfectly all right aswe will make the nozzle holder to suit. The important thing is to make all the nozzles in a set as near identical asyou can. If you are making several nozzles of different sizes,get all the machining done now whilst you are in the swing.

Tobe concluded

Left - Photograph 11: Lempor nozzle after milling both flats.

December 2011 ENGINEERING in MINIATURE I o•.•..•nn •..n"'~I.ft"+i.,.1W I~ •• D,.~£"W~.,...- BI~..-.•.,..:. •.•..\..• •..•••.t'''. ..,.U ••••~II"I.I~ \VI UUU~C. ~ la~""I.,t:J

by Bill Roebuck

Concluded from page 185, December 2011

With a set of 4 nozzles machined try putting them together to make a cluster. When you have retrieved them for the umpteenth time from the darker recesses of the workshop floor, take the easy way out and stick them upside down in a bit of btu-tack. You can then check with the callipers for the small and large diameters. I suspect it is only possible to specify these accurately in advance after some sort of complex trigonometric calculation process, but as they will vary anyway depending on the depth of the flats, it is probably best to take the easy way out and measure them. Photograph 12: View looking up/he nozzle Photograph 13:One used,one new. Note thicknesses Task 5: Make the Nozzle Holder cluster. of metal showing on nozzle ends. There isn't much to say about this. The picture in 'More Blastpipe Musings' article tightened to hold them - make sure the blast pipe and connect up. I fitted a small published in the January 2011 issue of ElM wire damps the nozzles below the point bracket to the nozzle holder to stop the showing a full size one being machined where they start to diverge. The nozzle blower ring being displaced during tube points the way here. The thread should cluster will then sit in the nozzle holder cleaning. match the one on the top of the exhaust ready for silver soldering. pipe from the cylinders, and the shallow Clean off the blu-tack, flux well, Task 6: Have a play recess diameter should approximately and silver solder with the nozzle holder Engines fitted with LemPor draughting equate to the larger diameter of the upright and the nozzle cluster held down take a bit of getting used to. The first thing nozzle cluster, with the hole through the with a small block of rusty iron. It doesn't you notice is the lack of noise - more a soft middle about the same size as the minor really matter if the iron wire gets soldered CHISS-CHISSthan a sharp CHUFF-CHUFF. diameter of the cluster. I found it best to as well - the main thing is to use enough Even when working hard the exhaust make the shallow recess a bit on the small solder to give a good fillet round and noise is much reduced, so if you are not size, then machine out 4 recesses to accept between the nozzles. I only had one total careful you can work the engine harder the nozzle major diameter as a snug fit. It failure out of six, and that was due to a than you might intend. The next thing you is easier to be accurate ifyou achieve these moment of stupidity with a big stick of notice is the fire. Gone is the white glare - recesses on the mill using the rotary table Easyflo-2. Whatever grade of silver solder instead the fire seems to only burn with a and a suitable size end mill (a good reason is used it must be capable of leaving a fillet bright red glow, and you start to wonder for using 'standard' sizes for the outside - one I tried was like water when hot and if it's about to go out. The third thing you diameter of the nozzles). It is also possible ran through the gaps. One other nozzle set notice is the steam. Lots of it. Netta now to file them out, but the problem here is to twisted during the brazing, but seems to just about steams against one get the larger radius only in the rim of the work just as wen in the engine. Probably with only the driver as a load, but with 4 nozzle holder- you need the recess to rest gives a nice swirl to the exhaust gases. adults behind the evaporation rate starts the cluster on. Clean up as required, and the finished to get embarrassing. The fire must get hot. Now for the difficult bit - getting the nozzles should look like photographs 12 as there always seems to be clinker after nozzle cluster to sit straight in the nozzle and 13 (the clip is to support the blower). a run - perhaps the greater throughput holder for silver soldering. The method I Fit the completed blast nozzle, of air at lower velocity gives more even adopted was to clamp the nozzle duster making sure the top of the nozzles finish combustion but less fierce burning. The in the vice, with the tips held more-or-less within your calculated distance of the one thing Ican be sure of is that the engine together by blu-tack. Thin soft iron wire base of the chimney choke. Usually this is will steam, so the object of the exercise was then wound round the nozzles and achieved anyway due to the height of the has been achieved. nozzle cluster. If necessary A by-product of the re-draughting Photograph 14: Relationship between blastpipe cap and make a small extension piece is the lack of 'robins'. My other engines blower. to fit below the nozzle holder with conventional blastplpes all steam to raise it a little. If your old well, but tend to throw out red hot bits if blast nozzle was the type asked to work a bit harder, as my collection with in-built blower holes of ventilated shirt sleeves to suit any you will have to make a new occasion bears witness. I had been driving blower ring as shown to go the rejuvenated Netta for a while before round the outside of the it dawned on me that the amount of half nozzle cluster (photograph burnt coal coming out was negligible. 14). The trick here is to only After a Sunday afternoon's public running part-drill the actual steam session the smokebox was often filled orifice in the jets, then they almost to the halfway line with ash without won't block up during the any noticeable fall off in steaming ability process of silver soldering (photograph 15). The greater total blast them to the tube. With the nozzle area seems to result in a gas flow three blower tips soldered through the chimney with insufficient onto the copper tube, it is velocity to pull ash along with it except for a simple job to drill gently very small particles, and these do not have until the drill breaks through the mass to travel any distance. into the copper tube. Bend into a ring to fit round your Continued on page 225 - ~ o..A.S + c.L{-.

ENGINEERING in MINIATURE January 2012