|
Post by KennLindeman on May 29, 2020 5:47:38 GMT
Trying to understand valve gears will sometime in a model engineers live cause quite a lot of headaches. But I am trying to get a better understanding of the Belgian's design as I have a 5' Ajax running, 5'Springbok being build and a 7 /4 being redrawn in 3D Cad I went to are club library and have taken our every book that deals will the valve gear as it s my mission to try and under stand this valve gear much better. I even found a book that dates back to 1920 describing the works of Walschaerts Valve Gear. IMG_20200529_061211 by Kenneth Lindeman, on Flickr IMG_20200529_061245 by Kenneth Lindeman, on Flickr
|
|
|
Post by KennLindeman on May 29, 2020 5:56:47 GMT
First thing I am trying to understand is the slide valve with its Lag and Lead. And how does the timing work
|
|
|
Post by KennLindeman on May 29, 2020 6:31:47 GMT
|
|
steam4ian
Elder Statesman
One good turn deserves another
Posts: 2,069
|
Post by steam4ian on May 29, 2020 8:10:17 GMT
In the words of Maria in "Sound of Music"you must "start at the very beginning".
Get yourself a grasp of simple eccentric and slide valve operation so you understand why having no lap means the eccentric must be set at 90deg to the crank with admission beginning at the start of the piston stroke and finishing at the far end of the stroke. Then understand how having steam lap allows you to advance the eccentric angle to get admission at the beginning of the stroke and with admission finishing sometime before the piston reaches the end of its stroke. Then consider why this might be desirable. Its then a quick lump to slip eccentric for reversing and then through to Stephenson's link motion. think about what happens to valve travel in mid gear.
Once you get a grasp of this, Walschearts gear becomes a means of mechanically adding two harmonic motions at 90 deg phase difference to describe an ellipse which provides a linear motion to the valve similar to that achieved by a simple eccentric.
Good luck.
|
|
stevep
Elder Statesman
Posts: 1,070
|
Post by stevep on May 29, 2020 8:48:18 GMT
In the words of Maria in "Sound of Music"you must "start at the very beginning". Get yourself a grasp of simple eccentric and slide valve operation so you understand why having no lap means the eccentric must be set at 90deg to the crank with admission beginning at the start of the piston stroke and finishing at the far end of the stroke. Then understand how having steam lap allows you to advance the eccentric angle to get admission at the beginning of the stroke and with admission finishing sometime before the piston reaches the end of its stroke. Then consider why this might be desirable. Its then a quick lump to slip eccentric for reversing and then through to Stephenson's link motion. think about what happens to valve travel in mid gear. Once you get a grasp of this, Walschearts gear becomes a means of mechanically adding two harmonic motions at 90 deg phase difference to describe an ellipse which provides a linear motion to the valve similar to that achieved by a simple eccentric. Good luck. I think this is the essence of understanding Walshearts gear. The two motions - one produced by the expansion link, via the return crank, and the other by the combination lever, can be analysed separately. With the valve gear in mid-gear, the expansion link imparts no movement, so everything is due to the combination lever. This movement is 2 x (lead + lap), and is fairly easy to work out, as you know the distances between the centre lines, and the stroke of the crosshead. (Only easy if you know what the lap and lead are, of course). The other motion is more complicated because (normally) the centre line of the motion and the line to the expansion link are at an angle. There is also the fly in the ointment of the inequality of swing of the expansion link, requiring the back-set for the connection of the eccentric rod to the link. But there are plenty of books, and simulators to enable you to design your own valve gear. As has been said in other threads, it is not so easy if you are following a published design, and are not made aware of all the design criteria. In that case, it is a question of 'blind faith', and building it as published. It will probably work, but not necessarily as well as it could have.
|
|
|
Post by KennLindeman on May 29, 2020 15:31:12 GMT
So we have had feedback for 2 of our steam gurus and I will have to read there comment couple times to make sure I understand there comments.
We have another chat "Locomotive Return Crank Setting For Walschaerts Valve Gear." which I am trying to follow and I wanted to link into there discussion but felt it may be a bit advance and though lets get the basics and can then you there discussion to add onto what we get to understand
|
|
|
Post by Roger on May 29, 2020 16:40:07 GMT
I'm certainly no expert on this, but I will offer a few conclusions from what I have understood. Contrary to Walschaerts gear being an improvement over Stephenson's from the point of view of Steam events, it's near universal use is due to the ease that it can be made and serviced. It's pretty horrible in my opinion, since it doesn't even provide equal valve events in forward and reverse. It's also a pain in the neck to design, with the change of one dimension causing a cascade of changes across the design. There's no simple way to adjust one key parameter as far as I can see, you have to massage and tweak things to find the least worse situation. It's also easy to end up with events that look good enough, but resulting in an impractical mechanical arrangement. This is the sort of mechanism that ought to be ideally suited for AI to design, since there are well defined goals, but no clear path how to achieve them. Don Ashton's book would be worth looking at, and so would playing with a simulator to see what effect changes will have. Don kindly sent me the model definitions so that I could view the indicator diagrams. That makes it much easier to understand what's going on. Don't forget that you're trying to design for a certain amount of maximum cutoff for starting, but also a progressive notching back in both directions. I've tried to explain what he passed on to me in this Wiki I started for SPEEDY. It shows the original LBSC gear, Don Young's variant, and Don Ashton's so you can see why the first two were deficient. To be honest, I've not studied any of this in much detail, it really doesn't interest me enough to delve any deeper.
|
|
uuu
Elder Statesman
your message here...
Posts: 2,800
|
Post by uuu on May 29, 2020 17:34:53 GMT
Looking at the different diagrams, I begin to see how to interpret them, and can see how Don Ashton's is best. Thank you.
Wilf
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on May 29, 2020 21:43:38 GMT
I recall explaining to the then IWMES Treasurer Norman Godrey (of blessed memory) some 30 years ago in The Railway Inn, Newport, Isle of Wight, how lead increases in Stephensons valve gear for his lovely steam launch; splitting a few beer mats and doing a few pics and diagrams on them with my pen after one monthly meeting.
I have a vague recollection of Norman buying me a pint after the penny had dropped and he had understood what I was explaining!
Some model engineers understand valve gears; others do not. Some with a bit of encouragement do!
|
|
|
Post by KennLindeman on Jun 3, 2020 17:25:22 GMT
Thanks to kind members of this forum I was about to get better quality drawings and inter the correct dimensions into Allan Wallace's Simulator. The cycle seems to be much better. Then John Hale send me some changes which Don Ashton recommenced for the Simplex which I think runs even better. I will try to make a movie of the sim and try post the link
Thanks for the help and input so far.
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Jun 3, 2020 19:31:35 GMT
Hi Ken,
These old books are really of very limited application. We had a shelf of similar in the IWMES bookcase cabinet from the days when apprentices in the IW shipyards had to prove they had some knowledge, but these old books are pretty useless. I learnt nothing from them.
You would be better buying from Camden a copy of Don Ashton's book.
Chas S Lake didn't understand valve gears; he just regurgitated other irrelevant stuff. Model Engineer magazine from the 1930s is littered with his rubbish!
I wouldn't waste my time reading your book, and getting terribly confused as a result!
Don't assume that a valve gear book written many many years ago like yours actually is relevant and informative; sadly it isn't.
Cheers,
Julian
|
|
|
Post by suctionhose on Jun 4, 2020 1:39:38 GMT
Don't know if this is at all helpful, however it is an observation of myself and the learning of VG design.
It seems almost instinctive - perhaps encouraged by pre Ashton writings - to begin at the wheel, carefully equalising angles of swing and symmetry, ultimately finding out what the valve will do.
This is the wrong place to start! Granted the motion begins at the wheel and the maths follows a course toward the valve. It is counter intuitive to think the opposite way but this is what one must do.
Don said something like "designing a gear to do what?" Personally, I noticed of myself, a paradigm shift from doing my best to apply "this must me so" instructions to thinking "the valve must be 'here' now - how can I make that happen?"
Once your mind is aligned with that objective, you are free to break so called 'rules', upset symmetry, distort angles of swing and do whatever it takes without restriction.
The case for achieving equal cutoffs at the expense of equal leads (up to a point) is also 'not what we've been taught' but entirely justified by wanting equal pressure & thermodynamic activity for each stroke, each way, in each cylinder.
I think Don underestimated what it would take to reverse what is a natural and logical train of thought. I can only speak for myself, but once the penny dropped as I describe above, I felt like I had the freedom to 'design' knowing what really mattered.
To understand a VG, first understand cutoff & lap. Understand lead, it's purpose in full-size and why it doesn't apply to miniatures in the same way. Understand amount of port opening is irrelevant - they are open or shut, nothing more.
From there you can learn the sources of valve movements and learn the box of tricks available to the designer to make the valve behave obediently.
Does this help? Maybe not. Having tried to emulate instructions by Greenly, Harris, Evans and many more, a feeling of competence only came after DA and not just from reading his book 50 times, many emails and cautious venturing of ideas later!
|
|
|
Post by KennLindeman on Jun 7, 2020 18:48:17 GMT
Is there anybody who understands the Walschaert and Allan Wallace's Simulator
I have drawn the Valve gear using dimensions given to me by John Vale which he got from Don Ashton for the 5'Simplex. It would be great if somebody could look at what I have done and comment please. I can mail you the data file
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Jun 7, 2020 21:08:07 GMT
Hi Ken,
What valve gear have you drawn and for which design? Valve gear is rather specific to a particular loco and gauge and whether piston valve or slide valve.
Cheers,
Julian
|
|
|
Post by KennLindeman on Jun 8, 2020 4:28:47 GMT
Its the 5' Super Simplex as per Martin Evans drawings (Walschaert inside) but I have used dimensions for the valve gear using info provide by Don Ashton
|
|
|
Post by KennLindeman on Jun 8, 2020 4:42:09 GMT
|
|
stevep
Elder Statesman
Posts: 1,070
|
Post by stevep on Jun 8, 2020 9:14:48 GMT
Ken,
It's rather hard to see on the video, but if those are the ports at the top of the round hole, then it looks like the valve is too long. In mid-gear, the valve should just open at each end (by the amount of lead that is required). This would normally be just a few thou. LBSC used to say "the thickness of a tram ticket".
|
|
|
Post by KennLindeman on Jun 8, 2020 17:32:45 GMT
|
|
|
Post by jcsteam on Jun 9, 2020 8:29:33 GMT
This thread is interesting, i will be looking at a friends Dixie loco in the coming months, as although a lot of issues are now resolved with it, and it is now running. The loco refuses to run slow, this may be a number of factors, but i have read that the valve gear wasn't designed correctly. So i am following this thread closely to hopefully be able to perform the calculations and determine what needs looking at, if anything regarding the valve gear.
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Jun 9, 2020 9:47:14 GMT
Hi Ken,
Your movie thingy is quite interesting as it shows what happens in mid gear and with no movement imparted to the valve via the return crank.
So, what you are seeing is the movement of the valve created only by the combination lever, which is 2 x (lap+lead). For expansive working in miniature and fullsize you need 'lap' on the valve; usually a minimum of 3/4 x the steam port width up to 1 x the port width, each end of the valve. The 'lead' element is more controversial; it is the amount by which the valve opens to steam on dead centres in full gear to give pre-admission. It does not tell you the angle of advance prior to the dead centres that the valve opens, and is a factor that needs checking IMHO - something drummed into me by Don Ashton.
Design considerations at a basic level in your clip would be checking that the combination lever proportions match the valve dimensions and what the lead and lap functions are. Note that the combination lever is in phase with the crosshead and hence piston, whereas the return crank is sort of 90 degrees out of phase. The rear of the valve rod may need to be adequately supported (I've had to correct bent valve rods on 2 Walschaerts locos), and that the bottom of the combination lever can come very close to the rear cylinder cover. Note that the anchor link front pin waves up and down during a cycle and this creates a non perfect movement to the combination lever during a cycle.
Note also that the eccentric rod is moving at differing speeds during the cycle due to the return crank's rotary motion being converted to a linear movement.
I had a hand in the construction of Martin Evans' 'Super Simplex' some 25 years ago. I don't recall any particular problems with the valve gear apart from working out the trunnion block centres, and bent radius rods, and those very complicated box type expansion links.
Cheers,
Julian
|
|