|
Post by Roger on May 14, 2019 8:52:43 GMT
Hi Roger. I can just imaging how you felt when you saw the actual alignment figures! You've made me think about checking my tailstock alignment too! That will be an interesting test as I had those 2 ex-Myford guys service the lathe only a couple of months ago. I don't have a problem as far as I know but it would be worth at least checking the concentricity of the bore to tailstock. Bob. Hi Bob, If be very surprised if many lathes are much better. For most things out makes very little difference, but when you're splitting hairs it helps if it's closer. It would be interesting to see how close other lathes are.
|
|
|
Post by Roger on May 14, 2019 9:01:52 GMT
Hi Roger, I refer to your statement: It's taken most of the day messing about with all of this, but it's been well worth it. It ought to make the injector cones easier to drill centrally when the time comes. Your ability to achieve so much in one day is amazing, it would have taken me much longer even if I had the knowledge to measure the tail stock height and the ability to improve it. I understand the need to have it as close to the lathe axis as possible especially when drilling very small holes, however in your set up you will no doubt include a pin chuck which will probably be held in the tail stock chuck. These two items will themselves introduce further albeit small errors, I was wondering if you are considering measuring in the final setup, the position of the small drill tip in relation to the lathe axis? Brian Hi Brian, I'm not a fan of pin chucks, they're generally rather crudely made and horribly out of true. I have a small chuck that holds really small drills, but of course PCB drills alli have a 1/8" shank. I made a simple chuck for those years ago which could be adjusted, I'll have to seer if I've still got it. It was made for use in the old Bridgeport we had when I needed to drill out some 0.2mm air jets and the runout of our chuck was more than the drill diameter! I think it's probably worth going down that route again since injector hole sizes are small. It needs to be clocked up very close to the position it's going to be used in though, you can't assume that anything is going to stay true if you move the design or wind the drill forward a long way in the tailstock. Assume nothing is my mantra in these situations.
|
|
|
Post by Roger on May 14, 2019 19:54:58 GMT
Out of interest, I've discovered that increasing the air gap between the magnets on the magnetic coupling has greatly reduced the friction in the whole arrangement and completely resolved the issue of the gears snagging from time to time.
The issue is that you need to have reasonable clearances in the bearings inside the water tank else it's likely to seize up if there are any dissolved solids that come out of solution. It's not a pristine environment even though it will be filtered.
The force attracting the magnets is roughly a square law with distance. If the air gaps are small, a slight movement of the mechanism that reduces the gap on one side and increases it on the opposing magnet will result in an out of balance force. This means that clearance in the output bevel gear bearing in my design is pulled to one side, causing friction against the bearing.
I've found that by moving the magnets further apart, the effect is much less marked and the mechanism turns much more freely. That's because not only are the forces smaller because the air gaps are bigger, but the small movement of the gear from size to side has less effect. In other words, if the play in the gear bearing is large, then close fitting magnets ie small air gaps, are going to have a large effect on the side load.
As it happens, there's still plenty of torque available to turn the indicator, so although it can't transmit as much, it's not required anyway. This is something of a breakthrough.
Designing these things to work reliably and with little friction is way more challenging than you might imagine!
|
|
|
Post by Roger on May 14, 2019 20:41:20 GMT
Don't look too closely at the 3D printed safety valve cover and shoulders, I really didn't expect them to print at all. That's because they're so thin and fragile and it had to be done with lots of supports which had to be broken away. Unsurprisingly, the front part broke off and I had to glue it back on. Anyway, the whole point of the exercise was to see if what I'd modelled will fit how the cladding ended up. Boiler making and cladding are more Analog than Digital, not precision Engineering. No matter how hard you try, a boiler is never going to be better than +/- 0.5 to 1.0mm in the main dimensions, so the precise outside of the clad boiler is not going to match the 3D model. However, it's pretty close, and I think the cover is close enough to be dressed to suit the cladding. It does look like the shoulders need to be a little longer, they don't quite reach the inside of the main cover. Anyway, it was an interesting exercise and could have saved a lot of heartache. You really don't want to have to make two of these! 20190514_212653 by Anne Froud, on Flickr
|
|
JonL
Elder Statesman
WWSME (Wiltshire)
Posts: 2,912
|
Post by JonL on May 15, 2019 10:27:16 GMT
Astonishing what can be achieved, such a complex shape.
How long will the magnetic field in your magnets retain their magnetism? Long enough that you won't need to adjust the air gap any time soon? I've no idea of the half life of modern magnets.
|
|
|
Post by Roger on May 15, 2019 11:44:21 GMT
Astonishing what can be achieved, such a complex shape. How long will the magnetic field in your magnets retain their magnetism? Long enough that you won't need to adjust the air gap any time soon? I've no idea of the half life of modern magnets. You would have to go to a smaller nozzle than the 0.4mm one I have to get a decent print of that, but it was good enough to demonstrate the fit. It would appear that the strength of Neodymium magnets reduces approximately 5% every 100 years, so I don't think we need to worry about that.
|
|
|
Post by Roger on May 15, 2019 21:49:38 GMT
Yet another iteration in the design of the water level gauge. This time I've extended the area on the arm that bears against the support arm to prevent the gears meshing too closely. That's not quite right still, so I'm going to modify the bracket this time because that print is a but scruffy. I also want to print it with 100% infill in case it leaks and wants to fill with water. 20190515_223803 by Anne Froud, on Flickr The guard on the outside is going to be moved slightly inwards too so that the gears can't move too far out of mesh. The retaining washer has also been increased in size. The hex pockets are for M3 Stainless Steel nuts which will hold on a Brass arm. PLA is too light for the job, I need something with a bit of weight to hold the arm down and provide the force to move the mechanism as the water level falls. I'm spending a lot of time on this to get it to work as freely and reliably as possible. I don't want to still be developing this on the finished locomotive. I'll be giving it a good test on the bench when it's all together. 20190515_223738 by Anne Froud, on Flickr
|
|
jma1009
Elder Statesman
Posts: 5,901
|
Post by jma1009 on May 15, 2019 22:09:34 GMT
Hi Roger,
I don't think I have missed any of your posts, but can you perhaps explain to a dunce like me how a float connected to gears with magnets with a try cock indicates the water level in the tanks please?
I have followed the making of the parts, but do not understand how it all works.
Cheers,
Julian
|
|
dscott
Elder Statesman
Posts: 2,438
|
Post by dscott on May 15, 2019 23:33:39 GMT
Part of the sitting round the table at Doncaster, was to assemble it from memory and even operate the gears and arm from the outer spindle, it is so strong in transferring the motion. The closest explanation is a magnet below a sheet of card and rotating the magnet, gets the nail or other metal item moving round. The test cock tap outer is to have a dot of paint on the I am EMPTY side!!!
Best regards David and Lily.
|
|
|
Post by ilvaporista on May 16, 2019 3:41:40 GMT
Hi Roger, I don't think I have missed any of your posts, but can you perhaps explain to a dunce like me how a float connected to gears with magnets with a try cock indicates the water level in the tanks please? I have followed the making of the parts, but do not understand how it all works. Cheers, Julian It's the rotary position of the tri cock handle that is important. As the float moves it rotates the handle giving you a direct read out of water level. As David said a red dot would give a low level warning.
|
|
|
Post by Roger on May 16, 2019 7:15:24 GMT
Hi Roger, I don't think I have missed any of your posts, but can you perhaps explain to a dunce like me how a float connected to gears with magnets with a try cock indicates the water level in the tanks please? I have followed the making of the parts, but do not understand how it all works. Cheers, Julian Hi julian, I probably didn't make that as clear as I might have done, I haven't shown the handle on the end. David and Adrian have it spot on. The 'try cock' isn't a valve at all, it's just a plain body with the handle being turned by the internal mechanism by means of the gear and magnetic coupling. The gear ratio was chosen to give a half turn on the handle from full to empty. The idea is to put a red dot on one end of the handle so that when it's visible you know it's nearly empty. Hopefully that makes more sense now.
|
|
|
Post by Cro on May 16, 2019 7:27:55 GMT
Hi Roger, I don't think I have missed any of your posts, but can you perhaps explain to a dunce like me how a float connected to gears with magnets with a try cock indicates the water level in the tanks please? I have followed the making of the parts, but do not understand how it all works. Cheers, Julian Hi julian, I probably didn't make that as clear as I might have done, I haven't shown the handle on the end. David and Adrian have it spot on. The 'try cock' isn't a valve at all, it's just a plain body with the handle being turned by the internal mechanism by means of the gear and magnetic coupling. The gear ratio was chosen to give a half turn on the handle from full to empty. The idea is to put a red dot on one end of the handle so that when it's visible you know it's nearly empty. Hopefully that makes more sense now. Out of interest, what stops someone turning the handle to the "wrong" position by accident and then you not knowing if its full or empty? Love the idea and mechanism though, awesome idea!
|
|
|
Post by Roger on May 16, 2019 8:45:13 GMT
Hi julian, I probably didn't make that as clear as I might have done, I haven't shown the handle on the end. David and Adrian have it spot on. The 'try cock' isn't a valve at all, it's just a plain body with the handle being turned by the internal mechanism by means of the gear and magnetic coupling. The gear ratio was chosen to give a half turn on the handle from full to empty. The idea is to put a red dot on one end of the handle so that when it's visible you know it's nearly empty. Hopefully that makes more sense now. Out of interest, what stops someone turning the handle to the "wrong" position by accident and then you not knowing if its full or empty? Love the idea and mechanism though, awesome idea! Hi Adam, You can't force it to stay in the wrong position, it springs back to where it's supposed to be. It only has one position too, it won't snap to any other position than the one where all the magnetic poles line up. You can't break it either because there's not enough force transmitted through the coupling to resist the applied torque. That would be a problem with a direct connection and a seal though.
|
|
|
Post by Cro on May 16, 2019 9:02:50 GMT
Makes sense
|
|
|
Post by Roger on May 17, 2019 22:22:43 GMT
I took the Axle Weighbridge to the club last week to help set up a locomotive where the springing was suspect, and frankly it's a pain to transport because the control box and the axle pedestals can't be separated. I originally had connectors for them, but the resistance of the connectors played havoc with the readings so it had to be soldered from end to end. That prompted me to buy this cheap and cheerful case which holds all of the parts needed, including the 3D printed supports for the bogies or fourth axle on 8 wheeled locomotives. I've changed the DC power connector to the right one, the pin in the middle was too small and it wasn't making a reliable connection. There's also an indicator LED added to the side of the box now so I can see when it's powered up. I should have done that in the first place. 20190517_142140 by Anne Froud, on Flickr I've also managed to shoehorn in an old smartphone and charger to display the results. The Bluetooth Electronics app uses the original power hungry Bluetooth which runs down a smartphone in under an hour, so it's as well to run it with a charger to make sure it lasts the course. 20190517_142220 by Anne Froud, on Flickr Back to the water level gauge... I've decided to use 1.6mm diameter Silicon Bronze Brazing wire for the handle and shaft. The originals look to be a Bronze of some kind. I do have an ER32 collet labelled 2mm-1mm but it won't even go as low as 1.6mm before the jaws become bound against each other. I'll have to see if I can source a better one with a more realistic range, say 2mm-1.5mm, which actually works. In the absence of a suitable collet, and my 3-jaw not holding it well, a split bush was needed so I could hold the rod for turning etc. Here I've drilled and reamed a piece of 10mm Mild Steel to 1.6mm and it's being slit with a 0.5mm slitting saw in one pass. 20190517_142454 by Anne Froud, on Flickr The starting end is spot on... 20190517_143037 by Anne Froud, on Flickr ... and the other is good enough. Multiple depth cuts would probably have been more sensible. The feed rate was 2mm/min 20190517_143049 by Anne Froud, on Flickr Here is't being used to hold the turned handle so I can drill a 0.6mm diameter hole half way through to locate the shaft. I've used the wobbler on both sides to find the middle. 20190517_164135 by Anne Froud, on Flickr Holding it is fiddly, so the Electronic 'Helping hands' was used. I'm keeping the parts attached to the stock else it's almost impossible to hold. 20190517_171013 by Anne Froud, on Flickr A minuscule shaving of 0.5mm diameter Silver Solder wire was added to the joint and covered with Flux then left to dry. 20190517_214508 by Anne Froud, on Flickr Warming it up dislodged the Silver Solder which had to be nudged back on top. It was heated very gingerly with a small hand held Propane torch attached to a can. I wanted to be certain that it didn't immediately glow Red hot and melt the whole thing! As it happens, I didn't see it turn Red at all, the Silver Solder just vanished into the joint. 20190517_214845 by Anne Froud, on Flickr I ran an M1.6 die down the end to make it look like it's screwed in, then reduced the end to the thread depth before finally turning the last 0.8mm down to 0.6mm for the location. Here it is assembled to the magnet holder. That's a scruffy 3D print, but good enough. Printing things to the nominal size ends up making nice press fits on the magnets and the shaft. 20190517_222812 by Anne Froud, on Flickr The shaft is a touch too long, the threads are standing out beyond the body. I'll fix the shortly. 20190517_223109 by Anne Froud, on Flickr Anyway, here's a wobbly video of it in operation. I might move the magnets out a little further on the bevel gear, I don't need the amount of torque it can transmit and it only makes for more friction. Holding the arm and turning the Test Cock feels really strange, like it's on a spring, and it snaps back into position when you let go. I'm now sure this is going to work just fine. 20190517_223759 by Anne Froud, on Flickr
|
|
|
Post by delaplume on May 17, 2019 23:07:55 GMT
Bloomin 'ek Roger --------- that's impressive and no mistake..
Good vid. but not long enough...can you show the gears actually meshing and working ??.........I didn't realise untill now just how small it all is...
You'll be going to Spec Savers soon at this rate....LoL !!.........I'm toying with the idea of a basic 3D Printer by way of an entry level but don't have the spons at the mo...
Good to see you and "The Lads" at the ME Show....and put faces to names...
Alan
|
|
|
Post by Roger on May 18, 2019 7:05:53 GMT
Bloomin 'ek Roger --------- that's impressive and no mistake.. Good vid. but not long enough...can you show the gears actually meshing and working ??.........I didn't realise untill now just how small it all is... You'll be going to Spec Savers soon at this rate....LoL !!.........I'm toying with the idea of a basic 3D Printer by way of an entry level but don't have the spons at the mo... Good to see you and "The Lads" at the ME Show....and put faces to names... Alan Hi Alan, Glad you like it. The handle was a bit of a challenge, that's only 6mm across the 'T' handle. I'll take another video when I've got the Brass arm attached. I've got a scrap part made from PLA immersed in water which has been there for two weeks to far. I'll be seeing if that changes in size or goes soft. So far it seems unaffected so that's promising.
|
|
|
Post by andyhigham on May 18, 2019 7:41:17 GMT
Are you going to fit working test cocks in the other two positions?
|
|
|
Post by 92220 on May 18, 2019 8:22:42 GMT
Hi Roger.
I've just watched your video. As Alan says, very impressive!! And a clever bit of 'outside-the-box' thinking to think it up and design it!!
Bob.
|
|
|
Post by Roger on May 18, 2019 8:38:28 GMT
Are you going to fit working test cocks in the other two positions? No, they are just going to be dummies. I don't really want water on the cab floor if I can avoid it.
|
|