|
Post by Roger on Oct 29, 2019 23:32:52 GMT
Here's the latest iteration of the shoulders for the Safety Valve bonnet with the end necked in. It's closer to the Works Drawings than what's fitted to 1501, but I think it's better that it's a close fit as was intended. The flange at the back is 0.5mm thick, but that goes down to 0.3mm thick at the bottom where it will go under the bonner skirt. 20191029_133527 by Timothy Froud, on Flickr They're handed, and this is on the wrong side, so it's not sitting quite right. 20191029_222316 by Timothy Froud, on Flickr 20191029_222331 by Timothy Froud, on Flickr I need a 7.5mm spanner for the miniature swaged fitting on the Top Feed, and I didn't fancy buying a whole set that includes one. Most of the sets look like they're stamped out of sheet, so that's not what I'm after. So rather than mess about trying to find something better, I thought I'd just make one from 4mm Gauge Plate using a 2mm cutter. 20191029_165651 by Timothy Froud, on Flickr 20191029_214846 by Timothy Froud, on Flickr The second top feed pipe was a lot easier to make, knowing the shape from the first one, even though it's a mirror image. 20191029_215303 by Timothy Froud, on Flickr 20191029_222707 by Timothy Froud, on Flickr I've cut the pipes down so the coupler bottom isn't seen below the bottom of the Pannier Tank. I need to set the angle so the flat roughly right so it will line up with the back of the tank where there's a flat inside the through tube. I've just set this by eye, I can bend the tube a little if it's not quite right. 20191029_230436 by Timothy Froud, on Flickr I thought better of playing the flame near the thread on the bottom of the couplers, so I shielded them. 20191029_230532 by Timothy Froud, on Flickr The flats of the top hex were painted with Tippex. I've used a ring of thicker Silver Solder to make sure there was enough for a nice fillet. 20191029_230824 by Timothy Froud, on Flickr Most of the Tipped came off in the Ultrasonic Tank, so there won't be much cleaning up to do. I'll try them in the tanks first though in case they need a bigger adjustment. 20191029_232050 by Timothy Froud, on Flickr
|
|
|
Post by jon38r80 on Oct 30, 2019 9:18:12 GMT
"I need a 7.5mm spanner for the miniature swaged fitting on the Top Feed, and I didn't fancy buying a whole set that includes one. Most of the sets look like they're stamped out of sheet, so that's not what I'm after."
Try looking for a set made by the Japanese firm "Engineer". They are not cheap but are die forged and are in metric/impewrial sizes from 3.2-1/8 to 11/32-9mm. or just make nice ones that you already have been doing. I dislike the stamped offerings available here intensely. I have had a set for years for working on RC cars but got another set last year when I was in Japan looking around a very well stocked model shop in Osaka.
|
|
|
Post by Roger on Oct 30, 2019 10:40:45 GMT
"I need a 7.5mm spanner for the miniature swaged fitting on the Top Feed, and I didn't fancy buying a whole set that includes one. Most of the sets look like they're stamped out of sheet, so that's not what I'm after." Try looking for a set made by the Japanese firm "Engineer". They are not cheap but are die forged and are in metric/impewrial sizes from 3.2-1/8 to 11/32-9mm. or just make nice ones that you already have been doing. I dislike the stamped offerings available here intensely. I have had a set for years for working on RC cars but got another set last year when I was in Japan looking around a very well stocked model shop in Osaka. Thanks for that Jon. I don't mind spending out of tools if they're decent quality. Narrow stamped ones are just going to leave nasty marks on Brass. That's why I ended up making it pretty chunky. I've made most of the box spanners, but there might still be a need for other small sizes.
|
|
|
Post by delaplume on Oct 30, 2019 12:14:19 GMT
Quote}-- "Gresley put things right in an instant: "quick, go and amend the drawings to specify what they have built!"
Ah yes---- that old "Horse and Cart" syndrome, eh ?? ---LoL !!..........
Seriously though, isn't that the hallmark of a good Engineer ?...The ability to think on your feet, be flexible and creative in your response to a changing environment ??........Well done Herbert.
|
|
Midland
Elder Statesman
Posts: 1,875
|
Post by Midland on Oct 30, 2019 13:33:40 GMT
Hi Roger I have yet to understand why seemingly sensible and mature people go weak in the knees over the GWR spin. Now we know Churchward was a good enginer, probably ahead of many and his "Saints" were a leap forward but we also know that not a lot of signficant development happened after that. Stanier discarded the swindon superheaters as too weak for example and why we should all have to go into gregorian chant about a bloody great ugly chunk of brass on top of a loco beggers rational thought. I was told form follows function in engineering, that is why the French locos are gorgeous, the GWR seemed to have grasped a picture and tried to make it work and ended up with funny looking tops where five millicockups excite! Bravo to Gresley, Johnson, Fowler, Stanier, Mansell et al, D
|
|
|
Post by andyhigham on Oct 30, 2019 14:46:56 GMT
The GWR did not like the locos workings to be on display, safety valves had to be hidden away and outside valve gear was resisted for as long as possible. Visible safety valves was akin to walking down the street with willy hanging out
|
|
|
Post by delaplume on Oct 30, 2019 16:49:07 GMT
David,
You appeared to have forgotten Churchward's Star class locos ( a pre-Titanic design incidentally ) and their long-term locomotive engineering influences....
Now this next bit may be the pot calling the kettle Black----( given 4930 Hagley's past performances )---- but let's keep to the thread's main subject shall we ??
Roger, as far as appearances goes that bonnet looks A-OK to me and will certainly add that certain "Swindon look" to yours, or anyone else's Speedy.....
Do please ignore all this "Betty Pickering" and crack-on with the rest of this VERY intriguing thread...
|
|
|
Post by silverfox on Oct 30, 2019 18:54:23 GMT
Just keep it going Roger, it looks lovely
|
|
|
Post by Roger on Oct 30, 2019 22:45:36 GMT
I've reworked the top end of the model to match the Works Drawings much more closely and ordered some more material for when I get a lull and can spare the three days it will take to machine another bonnet. In the meantime, I've been figuring out how to machine the shoulders, and that's been a challenge. I've only re-orientated models through 90 degree increments in the past, but that's not going to work for this one. I need to rotate the model through 40 degrees to be able to reach inside and out from all sides. That's involved creating new planes at 40 degrees and modelling an attachment stub to hold it to the rotary table. So this is how that looks on the model. The key things to note are that it's strong enough to support the cuts and that the part that has to be separated from the stub is inside the bonnet and not seen. Shoulder outside by Timothy Froud, on Flickr The 'U' shape is 2mm thick. Shoulder inside by Timothy Froud, on Flickr Here's the setup, basically the same as for the Bonnet, with a decent amount of material protruding to give enough clearance for the ER32 collet chuck. The bar is within a couple of microns of being round, so that's good enough for setting the tool height as long as it's been clocked dead true to the centre of rotation of the rotary table. Here the new 8mm cutter is being set to the radius of the bar so that the centre of rotation is zero in the Z-axis. This is vital because all of the machining operations assume that this is true. I just rotate the coordinate system through 90 degrees and create the tool paths with the new orientation. 20191030_211431 by Timothy Froud, on Flickr Here's a wobbly video to get some idea of the feed and speed of the roughing cut. I'm taking 1.5mm deep cuts at 50mm/min and it's virtually silent, much quieter that it is when you listen to the video with the microphone up close to the action! 20191030_222402 by Timothy Froud, on Flickr 20191030_230307 by Timothy Froud, on Flickr
|
|
|
Post by 92220 on Oct 31, 2019 9:53:27 GMT
Brilliant solution Roger!!! Can't wait to see the finish macgned article!
Bob.
|
|
|
Post by Roger on Oct 31, 2019 17:30:37 GMT
Thanks Bob. Ok, it's been an interesting and confusing day but some progress has been made. Roughing out the top... 20191031_090033 by Timothy Froud, on Flickr 20191031_091024 by Timothy Froud, on Flickr ... and the bottom went well... 20191031_105334 by Timothy Froud, on Flickr ... as did the pre-finishing from the top... 20191031_112145 by Timothy Froud, on Flickr 20191031_113740 by Timothy Froud, on Flickr 20191031_122754 by Timothy Froud, on Flickr ... but I could see there would be an issue with the sides. I decided that it would be better to machine these with the sides facing upwards because vertical surfaces are a problem when you want to get a radius at the bottom of them with a small diameter cutter. That's because the shank is larger than the cutter and that would foul the sides. So here I'm checking that I can just reach into the corner if I file off the top edge which needs to be trimmer anyway. This is a 2mm ball nosed cutter. 20191031_123608 by Timothy Froud, on Flickr Anyway, that does just clear... 20191031_124912 by Timothy Froud, on Flickr ... and resulted in an acceptable finish and a fairly small radius, albeit not really small enough. Some adjustment of that is going to be needed by hand and a radius added to the bonnet inside. 20191031_134611 by Timothy Froud, on Flickr The problems started rearing their heads on the reverse side, and this appears to be due to the software really taking offence at the geometry I've created. Why it's a problem on the back and not the front is anyone's guess. So on this cut, there was some gouging that I hadn't spotted although I could clearly see that the whole centre section was going to be missed entirely with the tool paths balooning out into space. I messed about with the parmeters no end but to no avail, so I decided to run it anyway and clean up the bit it missed by hand. Anyway, I stopped it at this point because I could see it was going to end up as scrap. 20191031_150756 by Timothy Froud, on Flickr A bit more experimentation with the parameters looked like I'd resolved the issue, but running it revealed it had introduced a new issue. Anyway, to cut a long story short, it would appear that if you try to follow the geometry too tightly (I usually use 1 micron and it's never been a problem) then where one facet of the model meet another, a tiny mathematical error can throw the tool path off at the edge. The solution is to increase the tolerances by which the path is allowed to deviate from the perfect geomerty, and it now looks like the path is going to work. 20191031_161312 by Timothy Froud, on Flickr 20191031_162428 by Timothy Froud, on Flickr So this is how the part cleaned up, and that looks fine other than the cockup. 20191031_164540 by Timothy Froud, on Flickr I've quickly ground away the inside where it fouls the Safety valve and offered it up to the boiler and this is definitely going to work. It's going to need quite a lot of careful adjustment to get it to sit nicely, but it's doable. 20191031_170259 by Timothy Froud, on Flickr So I'm just starting on the second attempt, which hopefully will be more successful!
|
|
|
Post by 92220 on Oct 31, 2019 19:08:42 GMT
WOW!!!!!
|
|
|
Post by Roger on Oct 31, 2019 19:31:29 GMT
Well, hopefully when I make a good one!
|
|
|
Post by delaplume on Oct 31, 2019 19:55:36 GMT
I've said it before-- and I'll say it again}--------- My gahst is absolutely flabbered !!
|
|
|
Post by Roger on Oct 31, 2019 21:03:14 GMT
A quick reminder of how the top feed pipes are connected inside the Pannier Tank through tube. The Brass connector is completely hidden, the top being about 30mm below the top of the tank and the bottom nut being about 5mm up from the bottom. I need to hold the top of the connector to stop it turning while the bottom nut is being unscrewed. It only has to be finger tight, the seal is on a radially compressed 'O' ring. 20191029_232050 by Timothy Froud, on Flickr The bottom nut is reached by this open box spanner... Bottom spanner by Timothy Froud, on Flickr ... and an extended version for the top. Fotunately, the top hex on the coupling will be in a position where it's roughly aligned with the way the tube curves out from the top. The bottom one could be at any angle. I'm expecting to only get that undone a flat at a time and then unscrew it by hand when enough of it is protruding from below the tank bottom. Top spanner by Timothy Froud, on Flickr
|
|
|
Post by David on Oct 31, 2019 21:30:33 GMT
It's a shame the first attempt didn't quite work but I'm guessing the next two will. CNC can certainly get you out of some tight corners when you reach a part and wonder how you'd make it by hand. I can't think of a way to do these other than pressing or casting.
|
|
|
Post by Roger on Oct 31, 2019 22:16:33 GMT
It's a shame the first attempt didn't quite work but I'm guessing the next two will. CNC can certainly get you out of some tight corners when you reach a part and wonder how you'd make it by hand. I can't think of a way to do these other than pressing or casting. That's life, CAM outputs can behave very strangely on occasion. I would have got away with the original cockup had it not been for me missing the other one. The whole part is absolutely covered with tool paths and unless you looked at it very carefully, there didn't seem to be anything wrong. I guess you'd have to machine and file a former and then beat Brass sheet around it bit by bit. It's certainly possible, but this suits me better. It's time consuming, but the results are worth it in my opinion. The full sided ones are indeed a pressing, it says so on the Works Drawing. Whether that was just an open shape without the flange or with one isn't clear. The taper on the boiler means that it won't fit properly without some hand finishing though since they're handed and had to fit different boiler types.
|
|
twombo
Seasoned Member
Posts: 120
|
Post by twombo on Oct 31, 2019 23:13:27 GMT
Just a thought, Roger.
The “cockup, as you refer to it. Seems to Correspond with the “mesh” lines in the rendering. I wouldn’t say “cockup ,so much as vagaries, in the application where things get a little ‘loose’ at times. kind of geologic fault line where one surface displaces unexpectedly.
I have run into this with Rhino, on occasion. I tried RhinoCam for a short time. I Never was able to achieve even, one tenth, your success.
Keep at it, you will tame the beast! Is it printing in 3d OK?
Bashing tin doesn’t appeal to me either. Explosive hydroforming? Now that might be fun. But where did I put that plastic explosive?
The way the cover fits the feed line is Really sweet.
Mick
|
|
|
Post by Roger on Nov 1, 2019 8:17:14 GMT
Just a thought, Roger. The “cockup, as you refer to it. Seems to Correspond with the “mesh” lines in the rendering. I wouldn’t say “cockup ,so much as vagaries, in the application where things get a little ‘loose’ at times. kind of geologic fault line where one surface displaces unexpectedly. I have run into this with Rhino, on occasion. I tried RhinoCam for a short time. I Never was able to achieve even, one tenth, your success. Keep at it, you will tame the beast! Is it printing in 3d OK? Bashing tin doesn’t appeal to me either. Explosive hydroforming? Now that might be fun. But where did I put that plastic explosive? The way the cover fits the feed line is Really sweet. Mick Hi Mick, You're absolutely right about the issues appearing along the edges of mesh lines. It's the first time I've been really caught out with this issue, although I've had more than enough of them when it comes to filleting complex shapes! I called it a cockup because in the end it's me in the driving seat and it's up to me whether to accept what it's produced or not. I'm now hyper aware of anywhere on the model where there's a region where the path disappears below the surface because that might not be just the odd micron. Zooming right in and even making the model transparent so I can see it more clearly can help. It's all a matter of experience and figuring out what to do when it all goes wrong. I've not had any issues with 3D printing in this regard, but that's not altogether surprising since the layer heights are huge compared to the machining tolerances. I imagine it doesn't have any difficulty in smoothing over these areas. I believe they used to form the ends of rocket fuel tanks with explosive forming of flat sheets into a hemispherical mould. Now that would be something to see. Hopefully the next ones will come out fine, I should have the first one done by the close of play today with a bit of luck.
|
|
|
Post by delaplume on Nov 1, 2019 12:25:37 GMT
Hi Roger,
Bearing in mind that the shoulders are always finished as painted ( ie --won't need polishing ) could you not fill in those slight "deviations" with something such as J B Weld ?? then hand-finish to profile ??...... Seems a shame to commit so much effort to the scrap bin...
There's a PM on it's way..
Alan
|
|