Thoughts on Pitman Drives for Sternwheelers

[Bierjunge]

Hi,  
 
I've made a few thoughts on how to design a kinematically satisfying electric sternwheel drive via pitman arms.  
My criteria were prototypical motion of the visible part of the pitman arms, simple and compact construction, and, if possible, homokinetic transmission. The latter means: The rotation of the drive shaft should be transferred 1:1 to the wheel shaft without cyclic variations in ratio, speed or torque oscillations, and as little play or backlash as possible, because play in links which act alternatingly pushing and pulling might cause jerks, noise and wear.
A few sketches of a handful of variants, some of which are well known, and some of which I haven't seen in reality yet, are attached.  
I would highly appreciate any comments, additions or experiences out of the praxis.
Is anyone aware of any other (satisfying) concepts?  
 
Regards, Moritz

[Eddy Matthews]

I wish I'd had those drawings before I started on the drive system for my sternwheeler Moritz!

I have eventually got it to work, but the pitman arms aren't driving the wheel - They do move realistically, but the drive is transfered via a chain and sprockets. So it's not the ideal solution

Having spent weeks getting it to work, I'm not going to re-do it right now, but maybe sometime in the future I'll use one of your solutions...

Thanks for sharing them with us...

[derekwarner_decoy]

Hi PD's .. & Moritz...not a criticism .. but back all those years ago  :gather they only had early LOG tables... no slide rules or pocket calculators or desk mounted  :computer PC's

My vote would be for your Variation 6.......but a true machining   of a quadrant arc was achieved over 300 years ago which would have facilitated the harmonic power & displacement resulant requirement as you note/question  - Derek

[gbritnell]

When I was building my boat I tried to come up with a realistic pitman drive system for it. I didn't have all the variations that you describe but did have a few. The ones that would seem to work the best are the ones that have a reverse pattern but as you stated they take up so much room in the hull. I actually did come up with a perfect example but it needed a cnc machine to cut it. I made a dimensional layout of the required action of the pitman arm in AutoCad. I then created a cam slot in an aluminum disc to replicate the dimensional points required to operate the pitman arms with no slip as you have shown with the Scotch Yoke variant. It worked fine except the diameter of the aluminum disc was such that you couldn't get it low enough to be in line with the center line of the paddlewheel. I couldn't design anything that would be simple for the average builder to construct without the pitfalls that you have depicted.
gbritnell

[mjt60a]

Many many years ago, I tried to make a mechanism that would operate a sternwheel in a realistic way (ie, as if the forward end of the pitman arms were connected to steam cylinders...) this is what I came up with, using bits of meccano and scrap brass. It didn't work very well, but that could be because the 'slides' (for want of a better description) are not in line with the centres of the two crank-thingies.......
...anyhow, similar to variation 5 but with the motor shaft even further forward..... :?
(edited for size of picture which I hadn't realised was the full size version  :?  )

[Bierjunge]

Mick, Thanks for the picture!

Many many years ago, I tried to make a mechanism that would operate a sternwheel in a realistic way (...) ...anyhow, similar to variation 5 but with the motor shaft even further forward..... :?

No, that's a perfect example of variant 2 (sorry, I didn't find a more logic classification).

The slight disadvantage I see of such arrangement with two mirrored crank drives facing each other, acting in opposite directions, is that the irregularities of the visible outer and the hidden inner crank drive ADD to each other.
I would guess that, if your mechanism did't stick, this was due to some play in the various joints and especially some elasticity, presumably due to the extremely short guided length of the slider.

Variant 5 is somehow different: The inner and the outer crank drive (preferrably with equal rod length) are both oriented in the same direction, so that the irregularities CANCEL each other instead of ADDING.
I would therefore guess that variant 5 should run much better, but I haven't tried it in hardware so far. We will see...

Regards, Moritz

[Bierjunge]

Sometimes you don't find the easiest solution in the first place...  
 
The Variant 4 (Scotch Yoke type) seems to be quite popular and has repeatedly been built, but has the disadvantage of being non-komokinetic (with the jerking and/or jamming problem described above).  
 
But why not turning it into a perfect homokinetic drive by making a curved slot or groove in which the motor crank pin runs? To make it clear, I've added one more sketch (I've called it Variant 4 1/2 because it combines the advantages of Var. 4 and 5).  
In this way, the length disturbances of the tilting pitman can be fully compensated.
 
But I've never seen it built so far. Why???  
Because the curved groove is more difficult to make? Because nobody else has thought of it so far?

[Bierjunge]

Almost five (sic) years later...

I've drawn a few free-hand sketches of how the Versions 5 and 6, favoured by me, might be realised:

[gbritnell]

Hi Moritz,
It's hard to believe this thread has popped up from so long ago. As regarding the curved link it would not be hard to make, for me, as I have a milling machine and rotary table for my hobbies. I make links like this for steam engine reverse mechanisms. The only problem would be if the link slot was not a constant radius. It would have to be plotted out to see what was needed.
gbritnell

[grumpysumpy]

hi guys
restoring an old sterndrive that has parallel coupling rods fitted. rebushed the links and rods to take the slop out of the system and now having trouble with jambing. i can trim the cranks and pivot bushes to get a smooth operation in one direction but jambs in the opposite direction. have tried all angles from 70 degrees to 110 degrees on the slave crank - no help. put a slot in one of the motor drive rods - no help. backing off by half a turn the screws that locate the pivot bushes reduces the current draw and gives a smoother operation in both directions. the links are identical and so are the rods.
is there a correct procedure for setting up this type of drive system

[gbritnell]

Way back when I did several layouts with my Autocad drawing program to see if I could come up with a viable intermediate link system for driving the paddlewheel. In regard to your drawings the first one would work only if the rod from the bellcrank was the same length as the rod from the crosshead to the wheel otherwise you would get some angular differences in the motion. As far as you curved link assembly I'm not sure how it would work without doing a mathematical layout.
gbritnell

[capnbob]

Hi Moritz,

I used two scotch yokes and , yes, there had to be some slop to prevent it from jamming but it seems to run smoothly and the sound seems similar to the popping of the steam engines.  But I really like your last two drawings.  with the curved slot.  Before I retired I worked with Solid Works.  I could draw all the pieces and put them together and rotate the shafts to check for jamming and smoothness operation.  That's what is needed here.  I wish I had Solid Works now.

Bob

[JL Frusha]

Doing my own research and found one.  Posted it in TowBoatJoe's old thread, too...



Driving the lower crankshaft works the connecting rods, pulling and pushing the Pitman, driving the rear crank.

[herrmill]

If there is anyone here who has the inclination, time & equipment to make a couple pairs of arms, I'd be interested in discussing further. I'm drafting plans now for several Ohio River sternwheelers that I'd prefer to build with pitman, not chain drives.

Let me know if anyone would be interested. Thanks.

[Bierjunge]

In Addition to the homokinetic solutions in the original posting and reply #6, here's even one more:



Moritz