3.5" Northumbrian Loco

[Baner]

Hi

It has been so long since I started this loco that I can't remember when I did. The original was serialised in Model Engineer beginning in 2009 so I'm guessing I begun maybe 2010. In that time there's been brief flashes of activity followed by years of idleness. I'd recently got money enough to buy some more of the castings and so set to work again.
Here's the progress so far:
Northumbrian by DAVID BANER, on Flickr

Northumbrian by DAVID BANER, on Flickr

Couple more:

Northumbrian by DAVID BANER, on Flickr

Northumbrian by DAVID BANER, on Flickr

Dave.

 

[Baner]

First little job was to finish off the front wheels. The crank pins were all that was left:

Northumbrian by DAVID BANER, on Flickr

Pretty simple parts.One end of piece of 5mm Silver steel was turned down to 4mm, the other end was turned down to 3.2mm and threaded 5BA.

Then the wheels were clamped to the mill table, orientated so that the drill would break through safely in the table slots. A piece of 8mm HSS was chucked and used to line up the axle hole with the spindle.

Northumbrian by DAVID BANER, on Flickr

The DRO was zeroed then the chuck moved 14.2mm across and the crank pin hole drilled and reamed 4mm.

Northumbrian by DAVID BANER, on Flickr

Finally the Crank Pins were Loctited home. (The stop pins on the eccentrics were loctited at the same time.)

Northumbrian by DAVID BANER, on Flickr

The couplings will be next and after that hopefully start to tackle some castings.

Dave.

 

[Baner]

The couplings are next. Enough 3/8" square stock is cut to length to make 2 couplings. It is faced off, centred and spot drilled on both sides:

Northumbrian by DAVID BANER, on Flickr

The piece is moved to the lathe and centered in the four jaw:

Northumbrian by DAVID BANER, on Flickr

It's then turned down to 4.7mm for 12.7mm and threaded 2Ba:

Northumbrian by DAVID BANER, on Flickr

The process is repeated on the other side to give this:

Northumbrian by DAVID BANER, on Flickr

Dave.

 

[Baner]

Continuing with the couplings. The material is then cut in half and clamped in the mill. The threaded section is standing on parallels whilst the square is butted up against the stop which is flush to the side of the vice. The square section is indicated to idiot check the set up:

Northumbrian by DAVID BANER, on Flickr

Part is then milled to finished height and a 3mm slot drill is run through, 4 passes to depth and 2 to finish sides. Light cuts are needed to stop the part being kicked out of the vice:

Grinding vice by DAVID BANER, on Flickr

Next is to drill a hole through the forks but a drawing review showed a minor problem:

Northumbrian by DAVID BANER, on Flickr

These 2 couplings are supposed to be joined together with a pin, but as drawn, (guessing at all the missing dimensions) the hole is in the wrong place, or the slot not deep enough or the radius is wrong or a bit of everything. I decided instead to reduce the width to 1/4":

Northumbrian by DAVID BANER, on Flickr

Then drill through 1/8" at 1/8" in from the edge. I then radius the part on the disc sander. Set up here is simple: a bit of scrap sheet steel is drilled through near the edge, a bit of round stock is inserted to use as a pivot and the part is turned back and forth by hand:

Northumbrian by DAVID BANER, on Flickr

Not the most accurate solution but very quick and good enough for this part:

Northumbrian by DAVID BANER, on Flickr

Dave.

 

[Baner]

Started work on the axle pump. I thought the casting would be a bit better than it was - getting two cylinders perpendicular doesn't seem very difficult...
Anyway, bit of eye balling got it clamped in the vice near enough, and flats were very lightly cut on the port face and chucking piece:

Northumbrian by DAVID BANER, on Flickr

The part was flipped onto parallels and the process repeated giving two parallel references:

Northumbrian by DAVID BANER, on Flickr

Flipped on it's side and clamped on the references a third larger flat is machined perpendicular to the last, then flip again and repeat:

Northumbrian by DAVID BANER, on Flickr

This gives a place for the vice to grip well enough to machine the ports.

Dave.

 

[Baner]

Now that I can grab the casting properly the ports are fairly simple. The port face is milled to finished height, and drilled through 3.8mm. Following that the port is drilled 5.5mm to about 8mm deep, and the top 4mm tapped 1/4" x 40:

Northumbrian by DAVID BANER, on Flickr

A stainless ball needs to seal on the lower face so a cutter is needed to form a convex surface. A simple D-bit is made for the task:

Northumbrian by DAVID BANER, on Flickr

The through hole is then reamed. Common practice to seat the ball is to hit it with a hammer, though I've heard pressing it in is better or maybe lapping. Does anyone have any experience or suggestions?

Moving on, the part is flipped and the other port receives similar treatment.
A side port is also required for the water outlet. The part clamped on its side and a piece of 4mm silver steel is inserted through the port. An edge finder is then touched off on the side of the rod to find the port centreline.

Northumbrian by DAVID BANER, on Flickr

The 4mm rod is then removed and the port face edge is found. The table is moved in 5mm and drilled through breaking into the port. Final operation is spot facing to provide a seat for a future bushing:

Northumbrian by DAVID BANER, on Flickr

Dave.

 

[Baner]

I hope I'm not posting to much detail about simple operations; I enjoy reading about set ups, so I'm tending to post about them too. That said I'd welcome any suggestions as to how I might have done things better.
Next job was the pump barrel . A center point was eye balled then a little dimple was made with a spotting drill. Then into the four jaw holding on the previous milled flats, the dimple is centered and the barrel lightly faced and center drilled. The tailstock center was brought up for support and the outside of the barrel trued up:

Northumbrian by DAVID BANER, on Flickr

The center was swapped for the fixed steady and the barrel faced to length:

Northumbrian by DAVID BANER, on Flickr

About the time of that photo I realized I didn't have a boring bar or measuring tools small enough to finish the bore. So I'll have to order a reamer to finish the job. Luckily everything is only rough turned so the part can be removed and something else done on the lathe while I wait.

Dave.

 

[Baner]

Some work on Pump parts:
Here's a quick drawing to help make sense of these parts:

Axle Pump by DAVID BANER, on Flickr

Part 1 is the top cap which seals the outlet and limits ball travel, an easy little job:

Northumbrian by DAVID BANER, on Flickr

Part 2 is one half of a fabricated elbow. A bit of 3/8 round stock is chucked and drilled 3.8mm at 12mm deep. Being a ball seat this is also reamed 4mm:

Northumbrian by DAVID BANER, on Flickr

It's then of to the mill and cross drilled 4mm breaking into the previous hole. Then an 8mm slot drill cuts a counter bore:

Northumbrian by DAVID BANER, on Flickr

back to the lathe and parted off:

Northumbrian by DAVID BANER, on Flickr

There are two of Part 3 which are soldered onto the inlet and outlet as per the drawing. The end 4mm is turned down to 6.35mm and tapped 1/4x40. About 3mm behind is turned to 8mm to fit in the counter bore made in the elbow. A 3mm hole is drilled through, then a parting tool is used to turn a small boss to register in the 4mm hole in the elbow. This little boss is probably not neccesary but it seemed like it would help seal if the solder didn't flow properly. Then the lot is parted off:

Northumbrian by DAVID BANER, on Flickr

Dave.

 

[Baner]

The pump body needed prepping for the outlet soldering. A bit of scrap was chucked in the lathe, a section turned to 6.35mm and threaded 1/4x40. One of the port ends was screwed on and the live center brought up to support the other side. The outside of the port was cleaned up, the pump body flipped and the process repeated for the other side:

Northumbrian by DAVID BANER, on Flickr

The pump body is then taken to the mill and outlet port counter bored with the 8mm slot drill same as the inlet elbow. Missed a photo of that.
The parts to be solder are fluxed up and installed ready for solder:

Northumbrian by DAVID BANER, on Flickr

Northumbrian by DAVID BANER, on Flickr

Parts are then soldered a dropped in the pickle:

Northumbrian by DAVID BANER, on Flickr

Dave.

 

[neil_1821]

Some great work here, keep it up.

With reference to seating the balls, I’ve always hit it with a hammer, discard of the ball that was hit for a new one and it should seal. Never had a problem.

 

[Baner]

Thanks Neil.

I've been researching ball seating over a bunch of forums. Plenty of people have the same experience as yourself. A lot have trouble with it too.
I'm writing up a post summarizing what I found. Stay tuned.

Dave.

 

[Cymro77]

Great work. I will be following. Thanks for sharing.

 

[Baner]

Thanks for looking in Cymro77.

Not a lot done on the loco this weekend, just one easy part, the axle pump ram:
Just a piece of 5/16" Stainless steel round cut and cleaned up to 26.2mm - whoops - drawing misread; make that 30.95mm:

Northumbrian by DAVID BANER, on Flickr

Then clamped in the collet block to mill a flat and cut a 2mm slot about 5mm deep:

Northumbrian by DAVID BANER, on Flickr

Then just drill through 3/32":

Northumbrian by DAVID BANER, on Flickr

Quick and easy:

Northumbrian by DAVID BANER, on Flickr

Shop time wasn't quite that dismal however, I also managed some lubricator pump bodies, half finished:

Lubricator by DAVID BANER, on Flickr

and put a new shelf up above the small lathe and mill for tooling:

Workshop by DAVID BANER, on Flickr

The old storage spot was the tool box under the lathe, but my two year old son has claimed it as his own and 're-organized' it all over the workshop.

Dave.

 

[Baner]

I've done a little research into ball valves so I thought I do a post on what I found.

For those who don't know ball type valves are found all over model locomotives. They're variously called 'check valves' or 'clack valves' and elsewhere 'one way valves' or 'non-return valves'. Here's a picture:

Ball and seat by DAVID BANER, on Flickr
Basically liquid comes in from the bottom,the ball lifts, liquid flows. If it tries to come back down the ball is pushed onto the seat and flow stops. This only works if the ball seats properly.

The traditional method for creating the ball seat is to drill the port, ream, drop the ball on top the hole and smack the ball with a hammer, the ball is thrown away and a new ball is used in service. The ball is typically made of Stainless Steel.

My own experience is fairly limited, but I vaguely knew that some people had problems with sealing. So looking about I found, (in typical internet fashion) that the traditional method works for about the same amount of people that it doesn't work for...Some problems are as follows:

-Not hitting the ball hard enough.
-Hitting the ball to hard.
-Not reaming the hole.
-Not hitting the ball square on.
-Not replacing the ball after hitting.
-Cheap, out of round balls.

And solutions:

-Press, rather than hit the balls.
-Ream the hole. The hole needs to be perfectly cylindrical to fully contact the ball. Also, the reamer helps debur the seat area.
-Use a close fitting drift. Machine something up that's a tight fit in the bore to make proper straight contact with the ball.
-There's a few options for balls to use. Good quality stainless will generally work but is a bit wasteful and there can be minor differences between hit ball and service ball. Some people use rubber balls,(O-ring material) but they can get stuck on the seat or sucked into the pipe. There is also ceramic balls, (silicon nitride being the most common) which have the advantage of being super hard. These can form the seat without having to be replaced. They are a little more expensive than stainless steel, however.

So what I took from all that, and what I plan to do myself, is use the vice to press in silicon nitride balls with a suitable drift.

Two other things to consider. One is ball/seat size. Fortunately this resource is available:

http://ibls.org/mediawiki/index.php?title=Design,_Care_and_Feeding_of_Check_Valves

Here is a table from the link:

Everett Clem Ball Seat Table1 by DAVID BANER, on Flickr

I'm unsure how the data was compiled but plenty of people claim it sound. So in the case of my axle pump, the seat is 4mm or 5/32 so I've got 7/32 balls on order.

The other consideration is the finishing of the seat area. A concave finish such as left by a drill is considered unsuitable. Flat is better, and some say convex is best, though I found nothing definitive other than don't drill finish. I made a D-bit style cutter for a convex finish.

That's about it for balls. There is also another type of 'T' valve sometimes used instead of balls, (think tap washers) but I think I've gone on enough. If you got this far well done. I'm not anywhere near an expert on this, I've just collected other peoples experiences into this post. If there is anything I've missed (or got wrong) please chime in.

Dave.

 

[neil_1821]

I’d be interested to see if you have any success with sealing using your method. I’m not sure you’ll be able to exert enough force in a vice unless you intend on using a hydraulic press. I think, as with anything it’s a matter of experience and what works for some may not work for you. It might be worth taking a bit of bar stock, drill it, team it etc and try various methods to see what works best

 

[Cymro77]

Dave - thanks for the info especially the sizing tables. They will come in handy.

 

[deverett]

Yet another way to make the ball seats:
Get a piece of bar that is a good sliding fit in the space above the seat.
Put a centre hole in the end of the bar and glue a ball in the centre hole. The ball is twiddled back and forth with a bit of pressure on the embryo seat and the seat is burnished - a bit like it traditional car valve grinding,. A little lapping compound may assist, but from memory was not mentioned.

Dave
The Emerald Isle

 

[Baner]

Neil - Thanks for the input. I had a few stainless 7/32 balls on hand so I did a comparison test of vice vs hammer, on a piece of 3/8 brass rod drilled and reamed 4mm. One end was pressed in the vice, the other hit with a hammer:

Ball seating by DAVID BANER, on Flickr

A suction test on either end revealed both methods sealed equally. Further inspection of the seats with a loupe revealed both seats to be almost identical in appearance. I'd speculate brass yields a certain amount and doesn't go much further with increased pressure. I'm going to further assume that hole preparation (reaming etc) may be more critical than the ball seating process.

At any rate, either approach seems to work. But as you said Neil - I think it's a good idea to test before committing to a method. I'll be testing the silicon nitride balls too, when they arrive.

Cymro77 - Glad you found info useful. Full credit goes to Everett Clem for the Seating Tables.

Dave - I'd heard of the lapping/burnishing method but never heard mention of how to hold the ball for that operation, so thanks for filling in the blanks.

Dave.

 

[Baner]

The pump continues:
I'd planned to ream the pump barrel to size, but for a similar price I decided to get something that would be used more than once:

Northumbrian by DAVID BANER, on Flickr

Small bore gauges are not super accurate but will get me close enough. Now I needed a small boring bar. I make these by cutting off a piece of HSS, then mill a flat on a length of silver steel (in this case 5mm). I cut a little piece of silver solder sheet, then stack all the bits up on and flux. Heat from underneath until the solder flashes, then grind to shape:

Northumbrian by DAVID BANER, on Flickr

So the pump body was chucked up again, and I'm back where I was 2 1/2 weeks ago. The bore is drilled, then slowly bored to size. Little boring bars like this are tricky due to deflection, but I got there in the end (and forgot a final photo):

Northumbrian by DAVID BANER, on Flickr

There's still more to go, this pump seems to be taking forever to finish...

Dave.

 

[Baner]

Next job is to turn down the barrel and thread it 1/2 x 32. I've got a Die to do the threading as the lathe is metric. I also managed to forget my lathe Die Holder is metric too. :wallbang:
More tooling! On the plus side the bigger lathe got some use for a change. An adaptor was turned up:

Northumbrian by DAVID BANER, on Flickr

And the barrel threaded:

Northumbrian by DAVID BANER, on Flickr

The pump ram is sealed by a gland that threads onto the end of the barrel. This part had been previous rough turned and threaded; it is now screwed onto the barrel and the bore finished to size, concentric with the barrel bore. Then off to the mill for some spanner flats:

Northumbrian by DAVID BANER, on Flickr

Dave.