Bilgram's valve diagram a graphical method to design a steam valve

[Dan Rowe]

I struck up an offline conversation with Chris about the valve gear for the Sabino because he made a comment that he wanted to make this engine with a lot of detail. The only valve gear I have really studied is Stephenson link motion. This is mainly because of my interest in Shay locomotives which are essentially a marine engine on a locomotive. The very first Shay built by Ephraim Shay was a marine engine on a flatcar. A Shay locomotive is really just a marine engine that got lost in the woods.

When I discovered Bilgram's diagram I was thrilled as I thought my struggle to learn valve gears was ended. Bilgram's diagram is a powerful tool to design a steam valve but it does not really help in the design of the rest of Stephenson link motion or any other type of reverse gear.

It was easy for Chris to get accurate measurements of the link motion but when I asked about the angle of advance he gave me an approximate number because it is not really possible to say looking at a built engine where the center of the eccentrics are in relationship to the crank pin for the cylinder. 

If the eccentrics were keyed to the shaft and fixed like a Shay and this was an IMPORTANT detail to the builder then the angle of advance would be a key piece of information. Lucky for me the valve travel and the angle of advance are listed on the drawing card index for every Shay.   

With those two numbers and knowing that Shays used 1/16" of lead, I can calculate the cutoff for that Shay. I do not think it is a coincidence that this is true.

I was not sure what would be a good number for the cutoff for this engine but in the book "Machine Drawing and Design" by William Ripper, there is a design for a two cylinder launch with all the parts and a Zeuner diagram. I thought I could simply look at the Zeuner diagram and state the angle of advance but it is really not quite that easy even if you have drawn that diagram as a learning example. I took the easy way out and drew a Bilgram diagram using ONLY the valve travel, the outside lap, and the lead. This worked to 74.92 degrees.

I also told Chris that for model engines it is customary to set the lead to zero. This is a match to the advice given by Dave on the Sabino thread.
                                                                                             

The last indicator card for Sabino that I saw was 75% cut off and just a smidge of lead.     I would dispense with the lead on an engine this small.   
Dave

More on how to use a Bilgram diagram to design a steam valve in the next post.
Cheers Dan

[crueby]

This should be an intersting journey, looking forward to the diagrams and methods!
Chris

[Admiral_dk]

A Shay locomotive is really just a marine engine that got lost in the woods.

       Thank you for that laugh Dan 

Per

[crueby]

A Shay locomotive is really just a marine engine that got lost in the woods.

       Thank you for that laugh Dan 

Per

Then a Lombard is just a train on portable tracks!

[Dan Rowe]

Per, I just really wish i was the one who came up with that phrase but It was Dave Latrope circa 2005 where I got it from.

Chris some Lombards like the Phoenix have vertical cylinders so a Shay on endless tracks.

As I mentioned I only used three numbers to calculate the cutoff for the Ripper launch engine.
1) Valve Travel
2) Steam or outside lap
3) Lead or the amount the valve is open at TDC

Most readers will know that TDC stands for top dead center and BDC stands for bottom dead center. This gets rid of the clumsy naming found in a lot of old steam books where the ends of the cylinder are called crank end and head end. A steam engine had a head at both ends. I am mostly interested in vertical engines so I draw valve diagrams with TDC at the top and BDC at the bottom.

Here is the Bilgram I drew to find the angle of advance of the Ripper launch engine.



Cheers Dan

[derekwarner]

Thanks Dan...I understand the geometry OK , but will be interested as to where some of the missing values are resultants of what?   

Derek

[crueby]

Also once you get farther along, can these diagrams predict what effects notching in the Stevenson link does?

[Dan Rowe]

Derek, I am scratching my head as to what you are asking. If you mean missing values for the Ripper launch engine they can all be found in the book. If you are talking about the Sabino the plan is to work the whole valve solution using the Ripper book as an example.

Cheers Dan

[derekwarner]

Thanks Dan....I had not heard of the Publication......this is my only find  Requium for the Little Ripper Launch Engine


https://books.google.com.au/books?id=6xNUEAAAQBAJ&pg=PT31&lpg=PT31&dq=Ripper+launch+engine&source=bl&ots=6Sc7K7QiEo&sig=ACfU3U2ufjRq5wHyU0JVIzFc2F9AXqkPhw&hl=en&sa=X&ved=2ahUKEwjWyduo1433AhWWzjgGHWjUAwoQ6AF6BAg8EAM

Derek

[Dan Rowe]

Derek, wow that is NOT the book I was talking about.
This is the one and the link should open to the Zeuner diagram for the engine.
https://archive.org/details/acourseinstruct00rippgoog/page/n184/mode/2up

Unfortunately, some vandal Ripped off the color plates of the engine which are the real gem of a very fine book on the art of drafting. (and I DO mean art)

Here is a link to ART prints of the work that shows even more views of the Ripper launch engine.
http://gouldstudios.com/me_SteamLaunch.html

Cheers Dan

[Dan Rowe]

Also once you get farther along, can these diagrams predict what effects notching in the Stevenson link does?

Chris, good question. The short answer is no in this thread we are designing the steam valve. It could be used for a single eccentric engine or any type of reversing gear.

To see what happens by notching back it would be best to use a program like Docksteader to see and compare the output curves.

Now when looking at the curves made by Docksteader's program how do you know when you change a variable if the change made the situation better or worse? It took a bit of thinking about that and the best I could come up with was to use the slip eccentric module for the same engine. A slip eccentric simply shifts the eccentric so the engine acts like a single eccentric engine and the curves look like a single eccentric solution. So my theory is if the change makes the reverse gear design look more like the slip eccentric curves then the situation is better and if the curves look less like the slip eccentric curve then you have made things worse.

The motor that drives Docksteader's program only goes in one direction so it is possible to turn a steam engine into a compressor by shifting the eccentrics far enough. I have done this and it is weird. (It is a program that is fun to throw a monkey wrench into)

The other way to see what happens when by notching back is to make a model. It was common to make a partial engine model to do just that in the days of steam. Now with 3D cad it is even easier to make a full model to see what is up with notching back with Stephenson link motion.

Now to get completely long-winded... The most common way to connect the eccentric rods for Stephenson link motion is with open rods. With open rods, the lead increases from full gear to mid gear. With crossed rods the lead decreases from full gear to mid gear. Open rods are the most common way to set up Stephenson gear. It is simple to change to crossed rods by simply shifting both eccentrics by 180 degrees. It would be interesting to check on what happens to the valve positions for both crossed and open rods to see what happens by flipping the eccentrics at several different angles of the crank.

Cheers Dan

[Dan Rowe]

I like to read about the career path of the engineering authors I have on my shelf or at least what it says in the front of the book. William Ripper was a Professor of Mechanical Engineering at the University of Sheffield.  The other book on my shelf by Ripper is "Steam-Engine Theory and Practice"
I have the fourth edition published in 1905. It is a very modern approach to the study of the steam engine that includes thermodynamics. I know that the last word made some readers think "not for me" but it is a good book and one of the best I have read on the subject and it is technical but not as hard to read as some later books on the thermodynamics of heat engines.

Cheers Dan

[crueby]

Also once you get farther along, can these diagrams predict what effects notching in the Stevenson link does?

Chris, good question. The short answer is no in this thread we are designing the steam valve. It could be used for a single eccentric engine or any type of reversing gear.

To see what happens by notching back it would be best to use a program like Docksteader to see and compare the output curves.

Now when looking at the curves made by Docksteader's program how do you know when you change a variable if the change made the situation better or worse? It took a bit of thinking about that and the best I could come up with was to use the slip eccentric module for the same engine. A slip eccentric simply shifts the eccentric so the engine acts like a single eccentric engine and the curves look like a single eccentric solution. So my theory is if the change makes the reverse gear design look more like the slip eccentric curves then the situation is better and if the curves look less like the slip eccentric curve then you have made things worse.

The motor that drives Docksteader's program only goes in one direction so it is possible to turn a steam engine into a compressor by shifting the eccentrics far enough. I have done this and it is weird. (It is a program that is fun to throw a monkey wrench into)

The other way to see what happens when by notching back is to make a model. It was common to make a partial engine model to do just that in the days of steam. Now with 3D cad it is even easier to make a full model to see what is up with notching back with Stephenson link motion.

Now to get completely long-winded... The most common way to connect the eccentric rods for Stephenson link motion is with open rods. With open rods, the lead increases from full gear to mid gear. With crossed rods the lead decreases from full gear to mid gear. Open rods are the most common way to set up Stephenson gear. It is simple to change to crossed rods by simply shifting both eccentrics by 180 degrees. It would be interesting to check on what happens to the valve positions for both crossed and open rods to see what happens by flipping the eccentrics at several different angles of the crank.

Cheers Dan

Do you have a picture or drawing showing the difference between open and closed rods? I've not heard those terms before.   

[Dan Rowe]

Chris, I can sketch one up it will be quicker than scanning Cecil H. Peabody who was the head of the Marine Engineering Department at MIT. If you have the book "Valvegears for Steam Engines" by Peabody turn to Plate XV in the back and read page 41 for the explanation. I can get this for the next post. The Sabino has a piston valve with inside admission and that is a new one for me and as you already know that flips the eccentrics just like switching to crossed rods.

With the suggested 75% cutoff and zero lead here is the preliminary Bilgram for the Sabino.



Notice that the angle of advance is now 30 degrees. This will be the case for any steam valve with 75% cutoff and zero lead.

I have made a physical analog of the Bilgram diagram in my head so I can easily make changes to see what is going to change with the valve.

I think the valve travel line is a pole. The steam lap is a tetherball. The angle of advance is the string supporting the ball. The lead is how much distance the ball is from the pole. With zero lead the ball is touching the pole. The Ripper launch is shown with 1/4" lead so that is the distance the surface of the ball is from the pole. For me, this works well so have a whack at the ball to see what happens.

Cheers Dan

[crueby]

I think I am seeing more of the relationships, check if I am right. Given the valve travel, if a certain cutoff is desired then the Bilgram could be drawn and the required lap and advance angle calculated? From that port size and valve sizes could be designed to match?