|
|
Dec 26, 2016 10:16:23 GMT
|
I'm not a toolmaker (or mechanical engineer) so please be gentle with me!
Could one piece cranks be made before CNC?
And in terms of your crank, if it was a conventional crank with counterweights it could be built in bits and assembled, but because of the shape that's not possible?
|
|
|
|
|
|
|
Dec 26, 2016 10:43:29 GMT
|
Yes,one piece cranks could be made pre Cnc.
The construction of a Rotary engine is such that it easy to add rotors and increase its capacity. However. Every new rotor you add ,needs a stationary gear to rotate around. The way you assemble a rotor motor with its seals etc,dictates that the maximum number of rotor journals you can have solid on a shaft is two. So when you start looking at multi rotor engines,you will see two fixed journals. In the case of a three rotor,the journal for the front rotor slides on,locating on a key and a taper. On a four rotor, rotor journals two and three are machined onto the shaft,with rotors one and four again,sliding on,keyed and tightening on a taper. Six rotor,? 1,2,5,6 slide on ,splined and taper with journals 3 and 4 machined onto the shaft
If you look back at that animated video posted a couple of days back,you can see when the shaft is put through the centre / intermediate plate,you can see it being wiggled to get it through. That's why it's a maximum of two fixed journals.Rotary engine does indeed have counterweights on the eccentric shaft,just they are fitted afterwards,once the rotor chambers are closed up,there is a counterweight either end. Front one just keyed. Rear one keyed and tapered. Rear one can be built into the flywheel ,or a separate item in the cause of an automatic transmission car. Aftermarket flywheels designed to bolt onto counterweights normally used for an auto,just unbolt the flex plate and bolt the flywheel on. I need to make new counterweights for the three rotor,as they need to be in different positions compared to a two rotor. Not difficult to do,just another type of machining I havent tried before. But how hard can it be? 😂
|
|
Last Edit: Dec 26, 2016 10:55:18 GMT by Deleted
|
|
stealthstylz
Club Retro Rides Member
Posts: 14,961
Club RR Member Number: 174
|
|
Dec 26, 2016 13:13:07 GMT
|
A mate of my managed to make his own conventional crankshaft. Even after your process he had to use the highly technical (and standard practice) way of straightening crankshafts - stick it between 2 centres, dial gauge to measure where the bend is then hit it with a lump hammer till its straight.
|
|
|
|
|
|
Dec 26, 2016 16:28:54 GMT
|
^^ Yep. You are lucky if they don't bend. All you can do is take all the precautions and find someone you can trust to do the final heat treatment.
|
|
|
|
|
|
Dec 26, 2016 17:23:02 GMT
|
JOHNNY BRAVO AND THE CARTESIAN PLANE Sounds like a title of a best seller. So. The Cartesian plane. It's used to find your way around on most Cnc machines, as a little bit left of centre doesn't really cut the mustard,when it comes to a definite position. Cnc machining per se isn't that difficult. A Cnc machine will only do exactly what you tell it to do. It's not open to interpretation,you say go there and off it goes. Basically you have 3 sets of axi , X Y and Z. On a milling machine, the X Axis is from left to right. Y is towards you and away from you,and the Z axis is down towards the machine bed and up and away from the bed. Most Cnc machines can be programmed in G code,something I got fairly good at in the beginning of my toolmaking days. My CADCAM software generates g code,so understanding exactly what is happening in a g code programme helps immensely. Two of the Cnc machines I use are equipped with Siemens controllers. Their Sinumerik system is a Windows based programming system that is very user friendly. However,one function I do like,is the ability to programme in Fanuc,or g code as well. So, seeing the shaft wouldn't fit the lathe I figured I could mill it out. The mill is normally configured as three axis, X,Y and Z, but it does have a removable forth axis that basically a controllable rotation device. This is the A axis. You then need to let the controller know what plane you are machining in. These are shown in the picture below, as G17 through to G19 So, having not don't any serious g code programming for a couple of years,as I have got more into offline programming,I thought a little cheat was in order. So I sent this to the programming support section of the tool supplier to see what their Siemens gurus would come up with And waited for a couple of days...
|
|
|
|
|
|
Dec 26, 2016 17:52:55 GMT
|
Got a response couple of days later along the line of "cannot be done. Would need a cadcam programme to write tool paths". Not what I had expected. Yes ,I did have a cadcam programme,but the 4 axis programming option wasn't something I had. Don't tell me can't be done. I sat down and started brushing up on. G code.... And so it began, the beginnings of a 4 axis milling programme to mill the journals on the shaft
|
|
|
|
|
|
Dec 26, 2016 18:02:23 GMT
|
JB....can you update the title to "every day is a school day" please?... christmas is long gone now!! Methinks the techies are in for a shock...are you sure that a crank will fit into a bodily cavity though?
|
|
'80 s1 924 turbo..hibernating '80 golf gli cabriolet...doing impression of a skip '97 pug 106 commuter...continuing cheapness making me smile!
firm believer in the k.i.s.s and f.i.s.h principles.
|
|
|
|
Dec 26, 2016 18:15:56 GMT
|
^^ Meh...Just started the 12 days of Christmas
" On the first day of Christmas, my true love gave to me, a healthy chunk of En40b". 🤗
|
|
|
|
|
|
Dec 26, 2016 20:29:27 GMT
|
Haha I know the feeling of sending something off to CADCAM support, only to be told it can't be done, then emailing them back a week later saying errr ok I've figured it out myself and no I'm not telling you how unless you give us a bigger discount next year :-)
|
|
|
|
|
|
|
I can fully understand the response I got,in that they are busy supplying a service to their customers who forget to put decimal places and feedrates into programmes and then blame the machine 😉 Anyway ,rather say "we don't know" ,rather than,"It can't be done". So, back to pulling my hair out..... For those who would like a bit more of an insight on gcode,here goes. Take a look at this programme ,that was just a test to see if I could get machining an arc in the right plane. G0 is a rapid movement,in a straight line. No federate required. G54 is the workoffset. Very important as it tells the computer exactly where your part is situated relative to a position it knows(G53) G53 is where the machine references to when you start up,so basically it's home address. G1 is a feed,and thus needs a federate added for it to work. In this case it's F500 which equates to 500mm per minute. This feedrate will stay set,so after the first use ,it will stay set,so no need to keep putting it in,unless you want to change it. G3 is also a feed rate,but in a circular motion,not a straight line. If you don't specify a feed,it will be the same as G1 I,J and K define the centre of the arc you want to mill using G3 S is a speed. Spindle speed,and in this case 1909rpm M3 denotes spindle roatation clockwise. M4 would be anti-clockwise. G17 and G19 describe the plane you are machining in. Most run of the mill stuff is in G17, but in this case we do change into G19 whilst using the 4th axis,because that's the plane we need to machine the arc in. G90 is absolute programming. So everything is relative from the zero position/ datum set in G54 G91swops into incremental programming,so any figures entered in will be from where the tool is now,not from the zero position. Very handy when changing planes. You can see we swop from G17 using G90 to using G91 when machining in the G19 plane A is a radial position measured in this case degrees. The 360 relates to the forth axis. So,once we have positioned the tool and want to start machining the arc,we now have told it,to now move the A axis 360 degrees. It's in incremental programming,G91, so the part will be rotated 360 degrees from where it is now, with a feed of 500mm per minute, as that was what was specified for both G1 and G3 So that's all luuurvely. And does it work? No. Bucking Fastard thing. Stupid machine.... Three days later I have it sussed. Schoolboy error. Remember me saying a Cnc machine only does exactly what it is told to do? Well that's exactly what it was doing. The position set for the centre of my part ,in G54 ,was X0,Y0,Z0,A0....So,with me saying to it move to A360, it did nothing. Why? Because A360 and A0 are the same place.It was just one of the little quirks of Cnc programmes that you sometimes come up against. With it being in incremental programming,G91, I would have thought it would have just moved 360 degrees from there,but it didn't. Quick cheat and use A359.999 and hey presto it mills the arc So the fundamental move for milling the arc is sorted out. Couple of days later a very composite programme running part Siemens Sinumerik with patches of g code plugged in,and it all works. Couldn't resist sending it back to tech support..... Right. That should have put the insomniac's to sleep 😂
|
|
Last Edit: Dec 27, 2016 3:11:04 GMT by Deleted
|
|
|
|
|
Keep it coming.
At the moment I am reading a rail line condition assessment report and flitting occasionally to RR.
Let me put it this way, Your post woke me up!
|
|
|
|
|
|
|
I can fully understand the response I got,in that they are busy supplying a service to their customers who forget to put decimal places and feedrates into programmes and then blame the machine 😉 Anyway ,rather say "we don't know" ,rather than,"It can't be done". So, back to pulling my hair out..... For those who would like a bit more of an insight on gcode,here goes. Take a look at this programme ,that was just a test to see if I could get machining an arc in the right plane. G0 is a rapid movement,in a straight line. No federate required. G54 is the workoffset. Very important as it tells the computer exactly where your part is situated relative to a position it knows(G53) G53 is where the machine references to when you start up,so basically it's home address. G1 is a feed,and thus needs a federate added for it to work. In this case it's F500 which equates to 500mm per minute. This feedrate will stay set,so after the first use ,it will stay set,so no need to keep putting it in,unless you want to change it. G3 is also a feed rate,but in a circular motion,not a straight line. If you don't specify a feed,it will be the same as G1 I,J and K define the centre of the arc you want to mill using G3 S is a speed. Spindle speed,and in this case 1909rpm M3 denotes spindle roatation clockwise. M4 would be anti-clockwise. G17 and G19 describe the plane you are machining in. Most run of the mill stuff is in G17, but in this case we do change into G19 whilst using the 4th axis,because that's the plane we need to machine the arc in. G90 is absolute programming. So everything is relative from the zero position/ datum set in G54 G91swops into incremental programming,so any figures entered in will be from where the tool is now,not from the zero position. Very handy when changing planes. You can see we swop from G17 using G90 to using G91 when machining in the G19 plane A is a radial position measured in this case degrees. The 360 relates to the forth axis. So,once we have positioned the tool and want to start machining the arc,we now have told it,to now move the A axis 360 degrees. It's in incremental programming,G91, so the part will be rotated 360 degrees from where it is now, with a feed of 500mm per minute, as that was what was specified for both G1 and G3 So that's all luuurvely. And does it work? No. Bucking Fastard thing. Stupid machine.... Three days later I have it sussed. Schoolboy error. Remember me saying a Cnc machine only does exactly what it is told to do? Well that's exactly what it was doing. The position set for the centre of my part ,in G54 ,was X0,Y0,Z0,A0....So,with me saying to it move to A360, it did nothing. Why? Because A360 and A0 are the same place.It was just one of the little quirks of Cnc programmes that you sometimes come up against. With it being in incremental programming,G91, I would have thought it would have just moved 360 degrees from there,but it didn't. Quick cheat and use A359.999 and hey presto it mills the arc So the fundamental move for milling the arc is sorted out. Couple of days later a very composite programme running part Siemens Sinumerik with patches of g code plugged in,and it all works. Couldn't resist sending it back to tech support..... Right. That should have put the insomniac's to sleep 😂 So basically it's like the BBC micro turtle that schools borrowed in the 1980s except instead of felt tip pen it's got a cutting device of some kind... Or a big trak... If you're a different vintage to me you might not know what I'm on about! Either way there's something deeply satisfying when you program machines using direct code like that and actually crack a problem like your 359.9999999r instead of 360.
|
|
|
|
Phil H
Posted a lot
Posts: 1,448
Club RR Member Number: 133
|
|
|
If it was an industrial robot I'd understand it a bit better - you force it to follow the configuration as they have various ways to reach the same point even if it chooses the way you don't want..
G codes are a mystery to me but probably because I've not had to use them although I keep coming up against 840D's and the like more often these days.
|
|
|
|
|
|
|
^^ Pretty much. 😂 Most Cnc. Machines today still run gcode. Most of the cadcam packages will output gcode too. Very often I write a program offline in Mastercam,then,seeing I have a basic understanding of the gcode it generates,I go in and edit bits to suit my needs. So, I am grateful for the time I spent right in the beginning on the big old mill in g code because it gave me an understanding of gcode many do not have...
|
|
|
|
|
|
|
If it was an industrial robot I'd understand it a bit better - you force it to follow the configuration as they have various ways to reach the same point even if it chooses the way you don't want.. G codes are a mystery to me but probably because I've not had to use them although I keep coming up against 840D's and the like more often these days. The Vmc and the small Cnc lathe have Siemens 828d controllers. I think it's a brilliant opt out,in that you can program in Sinumerik,G code,or a combination of both. When I really get stuck programming wise,it's normally my background in g code that sorts me out.
|
|
|
|
gatso
Part of things
Posts: 20
|
|
Dec 27, 2016 13:16:48 GMT
|
I've just been on line and learned double dutch ,,,,it was easier than trying to understand ,G..code ,,,ps still loving the thread tho
|
|
|
|
|
|
Dec 27, 2016 13:28:54 GMT
|
Whilst I was still trying to figure out why the program wouldn't run,(I hadn't figured out the 359,999 bit yet),I decided it would be worth sticking the 45kg lump in the old lathe and rough turn the ends down to 51,5mm. The spindle bore on the Cnc lathe is 52mm,so I figured it might just go in. There is always one. One doubting Thomas. A newbie at work who had never seen me turn on a conventional lathe. The others knew better not to place wagers with me,but he thought he knew better. £20 on me not being able to turn that lump down to half its weight during my lunch hour. Game on. Shavings flew. I stopped a couple of minutes before the end of lunch and went and got the scales.....
|
|
Last Edit: Dec 27, 2016 13:31:47 GMT by Deleted
|
|
Darkspeed
Club Retro Rides Member
Posts: 4,883
Club RR Member Number: 39
|
|
Dec 27, 2016 14:56:50 GMT
|
........ I stopped a couple of minutes before the end of lunch and went and got the scales..... Yep I have felt like that over Christmas as well - fecking mince pies.
|
|
|
|
|
|
Dec 27, 2016 15:55:45 GMT
|
|
|
Last Edit: Dec 27, 2016 16:03:03 GMT by Deleted
|
|
|
|
Dec 27, 2016 16:02:13 GMT
|
Shaft now fitted in Duly turned down some more And that's round about when it all fell into place as to how I was going to machine this shaft finished. Even though I had figured out the 4 axis milling programme,I ultimately didn't use it,as I came up with a better way. The programme was a help though,as you will see later.
|
|
|
|