Making Eliptical rod/tube?

[JasonBModerator @jasonb]

[JasonBModerator @jasonb]

For my next steam project I intend to make Anthony Mount's Easton & Anderson grasshopper beam engine but at double the size, the X braced frame that supports the end of the beam was originally a casting with eliptical section members.

Casting a one off is not really an option so I intend to fabricate and thought I would see if anyone had made eliptical section in the past, the size needs to be about 11/16×9/32 (11x7mm). I need 4No 10" lengths, the half circles at the ends will likely be easier turned as a ring. Material steel or brass.

I have a few ideas as follows

1. Reprofile some tube either by rolling or pressing, 3/8" dia would be about the right starting point.

2. Mill using a 1/4" roundover bit to knock the corners off and then blend by hand

3. Again on the mill but work out co-ordinates to produce the profile in a number of fine steps like the ARC function on a DRO and then final blending by hand

4. Buy it in, I have found a company that does what I want and will get a price on Monday but as they deal with race cars its likely to be an exotic cr-mo alloy and not cheap plus I would prefer to do it myself

So has anyone done something similar and if so how did you go about it?

 

Thanks, Jason

Edited By JasonB on 02/09/2012 18:39:07

[Jeff DaymanParticipant @jeffdayman43397]

Hi Jason,

You could use a corner rounding endmill as shown above to mill each edge of strips of stock. If the stock is passed through the endmill at an angle rather than at 90 degrees, an approximate ellipse will be cut. You will likely need a simple jig or fixture like a mitre box to hold the stock at the correct angle and also support it as each edge is cut. Three corners will need a shoe or screw adjuster to support and centre the stock for the last edge cut. A window for the endmill in two sides will be needed.

In the prototype shop of the first large firm I worked for in industry the proto shop foreman devised this method and I cut a lot of strips in the fixture he built – we were making a casting pattern and needed elliptical and semi elliptical sections (over 30 years ago now, but I have always thought the corner rounding endmill and angled stock was a smart idea on the foreman's part.

Might work for you.

JD

[Jeff DaymanParticipant @jeffdayman43397]

Forgot to mention one other idea for makiing elliptical tubing- if you have a set of bending rolls, a set of forming die rollers could be made to flatten car brake line steel tubing or brass tubing to an elliptical form. An elliptical recess would need to be turned into each of two rollers to make a true rolled elliptical section tube.

JD

[KWILParticipant @kwil]

You can now buy corner rounding insert cutters a well, you could also water jet cut the frame as one and then make the elliptical section by way of machining/filing.

[Neil GreenawayParticipant @neilgreenaway71611]

Hi there,

A number of years ago we had a repair to do on broken connecting rods on a packaging machine. The ends were simply turned from mild steel bar and brazed on. The original tubing was oval in section and had broken during a "crash" caused by a cam follower collapsing. We used a length of ERW tubing and flattened from round to oval in a flypress with the depth stop set to give the right l x h dimensions. This worked quite well and might be OK for your purposes – hand blending with files to finish the profile.

Neil

[blowlampParticipant @blowlamp]

Jeff Dayman's second answer sounds promising if you have some bending rolls.

You might be able to mill a semi-elliptic shape into some stock by setting the head over (about 39.5 degrees, I think) to get a 3.5mm radius from an 11mm diameter cutter and sinking the endmill up to its centre at the tip.

Two such pieces put face to face should give a reasonable elliptical hole. which could hopefully be used in a press to crush some tube to the required shape if done carefully and in stages.

 

Martin.

Edited By blowlamp on 02/09/2012 20:31:35

[JasonBModerator @jasonb]

Thanks for the answers so far

Jeff, I had not thought of the angular method, rather like passing timber over saw blade at an angle to make a cove, I don't have rolls but that was how I was thinking of making a roller them with shaped grooves.

Ken, The one that Bruce supply is a rectangular section casting that needs filing to shape just gets a lot to remove as the sizes increase, probably not to bad if I do a bit at a time.

Neil/Martin thats what I had in mind for pressing two plates with suitable half elipse recesses.

J

[David Clark 13Participant @davidclark13]

Hi There

Draw the section in cad.

Draw the section in cad say 4mm larger.

Draw horizontal lines 1/2 mm apart. Use the point where the horizontal lines cross the larger profile as the co-ordinates. Then use an 8mm ball nose cutter at each co-ordinate except the Z should be 4mm lower to allow for 1/2 the ball. Run along the bar, backwards and forwards will do, you are not taking a big enough cut to worry about climb milling. Turn the job over and do the other side. 

regards David

The profile below is round but illustrates the point.

Part of the ball profile is always in contact with the profile. 

The horizontal lines are 1mm apart but if I had drawn them at 1/2 mm it would have been to cluttered. You can also use this method in 3 D but it don't half take a long time to program it. Imagine a half round cylinder with another cylinder sticking out of it upright at right angles. Took me a couple of days but very enjoyable challenge.

Edited By David Clark 1 on 02/09/2012 20:35:25

Edited By David Clark 1 on 02/09/2012 20:39:44

[JasonBModerator @jasonb]

David thats what I was describing in item 3.

"3. Again on the mill but work out co-ordinates to produce the profile in a number of fine steps like the ARC function on a DRO and then final blending by hand"

 

J

Edited By JasonB on 02/09/2012 20:35:18

[David Clark 13Participant @davidclark13]

Yes but most people won't understand hence the drawing.

regards David

[JasonBModerator @jasonb]

Just an update, I have decided not to buy in eliptical tube so will be making the section myself.

I will still try the press/roll method but having just done the bearing supports using the co-ordinate method I think that will be the route I will go down.

These couple of shots will give an idea of whats obtainable just using the DRO to calculate a half circle, the elipse will need working out in CAD. and I will use a rounded corner milling cutter so the steps can be a bit bigger as it takes a while to work round a 1/2" dia in 0.010" steps!!

J

[Maurice Cox 1Participant @mauricecox1]

Concerning your grasshopper beam engine, I have an aditional photograph of this engine that I believe Anthony Mount did not have. He drew the swinging frame as having curved sides . They are in fact straight. I hope this is of some help. It wont be if you have already made it of course! I'm afraid there are a nubmer of discrepancies between Anthony Mount's model and the photos. These, and all of George Watkin's photos are held at The National Record, English Heritage. I think you would find it worthwhile to obtain copies.

[David Paterson 4Participant @davidpaterson4]

If I were doing this in Wood, it would be very simple. done between centres with interrupted cuts and two centre points at each end being the foci of the ellipse. round the final corners with a strip of emery. Ths is done all the time for 'freehand' oval cross sections, but if the centre points a re measured and the depth of cut controlled is pretty repeatable.

How would this go in metal?

david

[TerrydParticipant @terryd72465]

Hi Jason,

there appears to be an anomaly in your dimensions. I think that 11/16" = 17.5mm not 11mm as you say. Which would be the accurate dimension?

Regards

Terry

[JasonBModerator @jasonb]

Well spotted Terry, should have been 7/16 (11mm)

David turning between ctrs may give problems with flexing over the required length as they are quite thin not helped by the intermittant cut. I've used the same sort of method to turn the oval bosses on glands and it woks well.

Maurice, funny you should say that as I have two photos of other E&A grasshoppers which have the straight sides, I guess Anthony used a bit of artistic licence. The also don't have the disc crank which I'm still in two minds about though having made it out of a slice of 80mm diaCI it should help with the flywhel effect and help with slow revs.

J

[JasonBModerator @jasonb]

Just a quick update.

Thanks to Maurice I now have some photos of the original engine and it looks like the part is more oval than eliptical which makes things a bit easier. I was in B&Q the other day and picked up a length of their 10mm x 1.0mm ERW tube to have a play with.

After a few tests the best results were after annealing the tube to press it in the vice with a 9/32" drill in the jaws to give a constant to tighten down to, does not look too bad after a quick clean up.

It does come out of the vice with a slight dip in the middle of the flat section like a slight figure eight but a sanding disk soon blends it out.

 

There are some of the other parts that I have done on this engine in this thread

J

Edited By JasonB on 11/10/2012 18:49:12

[JasonBModerator @jasonb]

Well finally got this part of the engine finished. I made the four main parts out of the squashed tube shown above and the two half round pieces were machined from flat bar and then bent to shape. Used a bit of 5" PFC as a jig to keep things all in place while it was silver soldered, and it cleaned up quite well if I say so myself, click for larger images.

J

[Michael GilliganParticipant @michaelgilligan61133]

Looks very tidy, Jason

MichaelG.

[Jeff DaymanParticipant @jeffdayman43397]

The assembly looks great Jason, well done. A tricky job to be sure.

JD

[TerrydParticipant @terryd72465]

Hi Jason,

great work as usual, how did you manage to hold such a complex structure togather while brazing? I love your techniques shown in your postings on the other forum,

Best regards

Terry

[Bernard Laycock 1Participant @bernardlaycock1]

I was interested to see that it might be possible to turn the oval bolting flanges on glands. At the moment I do it by using small and large filing buttons and would be fascinated if someone could explain to me how they can be turned.

Bernard

[JasonBModerator @jasonb]

You will just have to wait until I get round to describing how that part was made over on MEM

What you can't easily see in the above two pictures is that I notched out the channel to take two rods

The split top bearing housing was clamped to one rod and the bottom "rings" held with grub screws to the other, some pieces of scrap copper tube were also used to stop things moving sideways.

The bearing housings also had small spigots machined onto them which located into the straight side tubes and some pins were added to do the same at the other end.

The other parts were all carefully scribed to the side rods which kept them quite well located and I just needed the two blocks of metal to hold the short diagonal pieces in place. To stop everything dropping the two firebricks were packed up so they supported all the parts.

It would have been nice to fillet braze all teh parts but not having oxy I opted for silversolder and a wipe of JB weld just to blend things together.

J

[JasonBModerator @jasonb]

Bernard, I'll have a look at my photos to see If I have any being done in the lathe but now I mostly use a boring head to cut the two larger curves. If Andrew Johnson sees this he may add the photos of some bosses he posted elsewhere last week.

J

[Sub MandrelParticipant @submandrel]

> a wipe of JB weld just to blend things together.

Miss!! Jason's cheating!

Neil

[Author]

Posts