Ian_C's workbench - P4 and S7 allsorts

Discussion in 'Workbenches, including workshop techniques.' started by Ian_C, 21 May 2017.

  1. Threadmark: "Obtain a running chassis" - part 5 - cylinders, motion bracket and connecting rods

    Ian_C Western Thunderer

    Some more assembling of parts made earlier and some more clearance challenges.

    While trial fitting the cylinders to the chassis the purpose of a couple of mystery parts on the etch was discovered. The etch actually had them labelled as cylinder spacers but I hadn't come across them in the instructions so had left them off pending enlightenment. They fit directly to the rear face of the cylinder sub assemblies and fill the gap between the chassis side plate and the cylinders. The value of that became obvious when fitted the cylinders to the chassis with screws for the first time. Without the spacer the chassis and cylinder backs bend when you attempt to tighten the screw. The spacer fills the gap and allows the screws be tightened properly.

    With cylinder assembly and motion bracket assembly fitted to the chassis they lined up very well and the ends of the slide bars needed only a very small adjustment to contact the pads on the motion bracket. I have noticed on the prototype that there are sometimes packing shims between slide bar ends and motion bracket. MOK 1 - Crewe 0. The drop link can be soldered to the cross head. The slide bars can then be soldered to the motion bracket. Note that when cylinders and motion bracket are joined this way the cross head becomes captive, the cross head nut not being able to pass behind the motion bracket, and if the connecting rod is permanently attached to the cross head then that also becomes part of the assembly. Make sure all the work on these parts is complete and that cross heads slide easily before joining cylinders to motion bracket.
    cyl brkt rod assy 1.jpg
    cyl brkt rod assy 2.jpg

    Anticipating a struggle for clearance behind the cross head I made sure the connecting rod..errr..connection did not project beyond the back of the cross head. The inside end of the rod was filed down a little to reduce the thickness and a small washer was turned up to go over the connecting pin cast into the back of the cross head and soldered on. The pin was filed down flush with the back of the cross head. On my cross head castings the rear flange that engages the inside edge of the slide bars was already quite thin so there was no opportunity to thin it further in pursuit of clearance. We're looking for every small bit of clearance we can gain, so worth looking to see if you can scrape a bit off the back if you're working in S7.
    cross head back 1.jpg

    With the assemblies complete they can be fitted back on the chassis and the connecting rods put on the driving crankpins. More clearance problems, but not where I'd expected them. There was a sliver of daylight between the leading crankpin nut and the back of the cross head with the leading axle over at maximum side play. Same on both sides and somewhat of a relief. Reminder - to get this result in S7 you need:
    • minimum sideplay on leading axle, there'll always be some but as little as you can engineer
    • reduce the projection of the driving wheel bosses to scale thickness, Slaters wheels being fatter than scale in this respect (earlier post)
    • counterbore the leading coupling rod and recess the leading crankpin nut, you'll need to be more or less flush with the face of the rod (earlier post)
    Thank you to those who have posted useful observations on these issues on previous MOK 8F threads. You helped me to get it right first time! Good old WT eh?

    The real clearance issue was between connecting rod and the coupling rod joint between axles 1 and 2. It turns out that the side play I'd provided on axles 2 and 3 allows the connecting rod to move across enough to strike the coupling rod joint nut.
    rod clearance 1.jpg

    Having run the chassis in a little prior to this with just the coupling rods fitted I thought I'd accurately measure the axle side play I actually ended up with as opposed to what I'd calculated I'd end up with (earlier post). Turned out that I had more side play on all axles than I'd aimed for, and that presented the opportunity to reduce the side play on axles 2 and 3 slightly. Back to the CAD drawing board and taking advantage of the actual side play on axles 1 and 4 (that I'd assumed to be zero for the purposes of constructing the geometry originally) I found that I could put spacers behind the wheels on axles 2 and 3. For what it's worth I ended up with 0.16mm spacers on axle 2 and 0.44mm spacers on axle 3. Taken together that still gives enough accommodation to negotiate a 6 ft CL radius curve with a tiny bit to spare (in theory anyway - yet to be proven!), and just gave some clearance behind the connecting rod to miss the coupling rod joint. A benefit of having a lathe is that it's the work of a few minutes to turn up spacers of any size you require, and yes it is possible to part off spacers 0.16mm thick.

    And here's Mr Stanier's 0-8-0. Look's like I'm getting somewhere.
    con rods assembled 1.jpg
    One observation, at maximum piston stroke the end of the cross head just leaves the parallel portion of the slide bars adjacent to the motion bracket. The motion drawing on page 54 of the wild Swan book shows that the prototype does a bit of this too!
    Last edited: 3 March 2018
  2. adrian

    adrian Flying Squad

    I came across the same problem when making the motion on my Ivatt 2 - my solution was to make a little dovetail joint between the slide bars and motion bracket - it worked remarkably well (just a shame it was started in finescale :( ).
    more details here http://www.cherryclan.com/2015/08/ivatt-slide-bars/
  3. Threadmark: "Obtain a running chassis" - interlude - rolling road

    Ian_C Western Thunderer

    Obtaining the running chassis by driving it up and down on 3 ft of track wears thin after a while. A rolling road's what I need. So I bought a relatively low rent version from http://www.directtrainspares-burnley.co.uk . It was for O gauge 32mm and predictably S7 wasn't going to sit neatly on the rollers. Easy enough to modify though.

    The holes in the black plastic base plates were slotted out to allow the angle sections to be moved further apart. The ball bearings are a 'value' type and there's a bit of free play on the outer race. That's not a big problem in itself except the bearing inner races are bolted directly to the inside faces of the angle sections. That allows the outer race to rub against the angle under typical running conditions. 0.5mm spacers were turned to space the bearing off the angle (could probably have used M3 flat washers if I'd had any). The distance between the angles required to set the bearings at 33mm was calculated and two slots were milled in a strip of Tufnol. The rollers at one end are bolted to the Tufnol, the other sets are free to slide along the slots. The sliding rollers seem to find their own position under the wheels when running and seem happy to stay there without the need for bolts or clamps.

    rolling road 1.jpg

    Set up the loco on the rollers, applied power and off it went. But all wasn't well. Even after preventing the bearing outer race from rubbing on the angle sections there was still enough rolling resistance from the 16 bearings to cause the loco to want to climb up the rollers. It reached a perilous equilibrium, occasionally making little lunges for freedom over the leading rollers. Adding some weight to the chassis to make it sit down on the rollers caused the motor to run slower and you could hear that the drivetrain was working quite hard. The bearings were a type with a metal shield secured in the outer race. The wobbliness of the outer races caused the shield to rub on the inner races on most of the bearings. the remedy was to pry out the 32 shields with a small screwdriver. That freed things up considerably and the loco then settled down to run happily.

    rolling road with chassis 1.jpg

    In the end it's a simple device and I could probably have come up with a better solution myself for less. Knowing what I now know I'd have spent more money on a better quality of bearing. Anyway it works.

    One observation. The rolling road doesn't half amplify the wobbles. Where it seems to glide smoothly up and down the track it shuffles and shimmys on the rollers. Not totally unlike an 8F overdue for shopping trying to run at pace. Pity the Modelu crew.
    Rob Pulham and Len Cattley like this.
  4. Threadmark: Posting problems tonight?

    Ian_C Western Thunderer

    Forbidden - "You don't have permission to access /index.php on this server."
  5. Threadmark: “Obtain a running chassis” - part 6 - return crank and eccentric rod revisited

    Ian_C Western Thunderer

    A few posts ago I thought myself very clever having made a ball bearing big end for the eccentric rod. It was secured to the return crank with a small nut on the inside of the crank. A WT’er with more experience and sense than me pointed out that with a limited clearance between the crank arm and the connecting rod it was unlikely to work. That prediction came true! Back to the drawing board to rethink that one. In the end a small amount of cheating was required. Effectively I’ve turned the assembly back to front and used the bearing cover as the securing nut. I also took the opportunity to improve the appearance of the crank arm by representing the four counterbored holes present on the prototype. A picture should make the arrangement clear.

    assy explode 2.JPG

    assy explode 1.JPG
    Today we have naming of parts…

    1 - The bearing cover unsoldered from the original assembly.
    2 - A turned flange and bush. Threaded 12 BA internally and soldered to the bearing cover.
    3 - The eccentric rod.
    4 - A 12 BA stud soldered into the crank arm.
    5 - The outer layer of the crank arm.
    6 - The inner layer of the crank arm.
    7 - Another turned thingumabob. Soldered to the inside face of the crank arm and with a 10 BA stud to screw into the driving crankpin.

    Not shown - four short lengths of wire to represent the prototype studs and nuts. Soldered into the holes in the inner layer of the crank arm.

    The crank arm layers are made from 0.5mm NS sheet. Drilling holes tidily in thin sheet metal can be a challenge so the 0.5mm strip was soldered to a 0.8mm strip of scrap NS to make the job easier. The whole was screwed to an offcut of timber and clamped to the milling machine bed.
    drilling crank holes.JPG

    All the holes in the crank layers were co-ordinate drilled and the blanks cut out and unsoldered from the backing strip. A small tool was turned to help with the marking out of the crank profile. It’s just a rod with a diameter and spigot each end. The diameter of each end matches the diameter at one end of the crank.
    crank 2.jpg

    The spigot is located in the appropriate hole and the rod is scribed around. Something I’ve learned - a thin coating of solder holds the scribed lines much better than marking blue or a spirit marker pen, particularly when there’s a lot of filing and handling involved.
    marking out crank.jpg

    The parts soldered together and ready for assembly. Well, the bearing cover hasn't been soldered to the flanged nut thingy yet. The hole in the eccentric rod big end was opened out to be a clearance for the threaded bush.
    crank and rod parts.jpg

    The finished assembly.
    crank and rod assy.jpg

    The crank arm is flush on the back and clears the connecting rod, and the bearing cover/nut is easy to remove to dismantle the motion. The cheat is that the bearing cover rotates with the crank arm now rather than being fixed to the eccentric rod. In practice this isn’t very noticeable when the loco is in motion. That gets me on the way to assembling the rest of the valve gear.
    Len Cattley, Arun, Rob Pulham and 5 others like this.
  6. mickoo

    mickoo Western Thunderer

    There is an alternative way, but again requires a little cheat.

    Get rid of the bush (2), solder the bearing cover (1) to the eccentric rod (3), solder the pin (4) to the eccentric rod (3), pass the pin through 5 and 6 and solder a small thin washer to the pin behind part 6.

    The cover will now stay aligned to the rod and the whole lot rotate freely; whilst the pin is not threaded, once the washer is on it's on.

    You may have already thought of this, in which case I've misread your words ;)

    Nice work either way :thumbs:
  7. Ian_C

    Ian_C Western Thunderer

    I have so little clearance behind 6 in the worst side play condition that I decided it had to be flush. Besides I wanted to be able to separate the assembly at this point. Good solution where there's enough room though.
  8. Threadmark: Connecting rod cinema

    Ian_C Western Thunderer

    Gather a small amount of popcorn...
    Rob Pulham, P A D, michl080 and 2 others like this.
  9. Threadmark: “Obtain a running chassis” - part 7 - valve gear pins

    Ian_C Western Thunderer

    A lot of the smaller motion pins on the prototype were secured with collar and a taper pin. They're small features on the model but it's worth trying to reproduce them. A distinct improvement over seeing BA nuts or screw heads holding the valve gear together. There are a number of small motion pins that are very similar in size so I've standardised on a 0.9mm pin for all of them. Start by turning some collars, in this case 1.4mm diameter, 0.5mm thick with a 0.9mm hole through. You need to be spot on with the parting off tool on items this small, but once set up it's dead easy to knock off a handful. You'll lose a few so make some spares!

    Solder the collars to short lengths of 0.9mm wire. Put the pins in your Proxxon or Dremel mini drill and take the end of the pin nearly back to just proud of the collar with a file. Polish up what's left with an abrasive block, like Garryflex, while still in the drill.
    valve gear pins.jpg

    Fit the motion together with the pins. Cut off and dress flush on the rear of the joint. Secure with a tiny flash of solder. If the hair shirt of authenticity really itches you could try and drill through the collar 0.3mm and fit a length of wire to represent the taper pin. That's Monday evening's quota of gumption all used up.
  10. Threadmark: "Obtain a rolling chassis" - part 8 - valve gear complete

    Ian_C Western Thunderer

    It took a lot a assembly, disassembly, adjusting and tweaking to get all the clearances right and all the parts moving without restriction, but finally the valve gear is complete on one side. I originally said I'd set the gear in neutral but decided I'd prefer to see some valve motion so I set the radius rod at about half forward gear in the expansion link. I set the gear at full forward cut off on the other side, just for the hell of it!

    Clearances were tight again, no surprises there. Getting the connecting rods to pass behind the lower part of the expansion link required some fettling. I had to file off part of the overlay on the inside face of the expansion link assembly to create clearance. If I'd known I would have just snipped off a bit of the overlay before soldering it on. Just make sure you do one of each hand!
    expansion link 1.jpg expansion link 2.jpg

    The combination lever is straight as assembled from the etched parts. If you leave it straight then there will be very little or no clearance behind the lever to the leading edge of the crosshead and it will tend to pull the front end of the union link out of line. The motion drawing on pg 54 of the Wild Swan book (easily the most referred to page in the book so far on this build, and, in the absence of detail instructions, invaluable) shows that it is slightly cranked at both ends on the prototype to bring the lower end out in line with the union link, and that provides just enough clearance for the cross head. Easy enough to do on the model part. There's an oil reservoir at the lower end of the combination lever that is not represented on the model. It isn't do-able on a flat etched part so I can understand why it's not there. I agonised for a few days and decided not to try and add it.

    The rotation of the combination lever in the valve crosshead casting needs careful attention. With the parts as supplied there's simply not enough room for the combination lever to rotate. The further you set the gear from the neutral position the more rotation you'll need at that pivot. It's relatively straightforward to open out the casting with a needle file or piercing saw.

    On the prototype the radius rod is free to slide in the expansion link, it's position in the link determined by the position of the reverser lifting lever. On the model the rod is pinned to a position on the expansion link. Kinematically then it it will conflict with the fixed position of the lifting lever when the motion is...umm...in motion. Only a little, but we want the valve gear to move freely so better to avoid the clash. Therefore I didn't pin the lifting lever through the slot in the rear end of the radius rod. I added a dummy pivot pin end to the lever and left the rear of the radius rod free between the plates of the lever. The rod appears to slide back and forth in the lever and the compromise isn't detectable.

    One last problem to overcome was the lack of restraint of the valve crosshead in its carrier. The radius rod is very light and it only takes a small amount of friction at the rod to expansion link pivot to cause the radius rod to lift on part of the stroke, and in so doing it tends to lift the combination lever a little. It doesn't stop the gear from working but you see a little hiccup in the motion when it happens and that spoils the illusion. The solution to that was to solder on a restraining bracket made from a tiny length of brass angle filed to the right height. It's added to the inside of the carrier, behind the combination lever and radius rod, so hopefully not noticeable in the murk beneath the footplate. A photo should make that clear.
    piston valve crosshead 1.jpg

    Eventually, working valve gear!
    valve gear 1.jpg
    combi lever and crosshead 1.jpg

    And, naturally, there's video...
    Last edited: 17 March 2018 at 11:52
  11. Ian@StEnochs

    Ian@StEnochs Western Thunderer