www.Robse.dk - Thé internet site!!

Projects by Robert Holsting

Blueprints & plans
About me

  Cardans, steering, gearboxes, main motor
Once again everything begins on the whiteboard. It soon became obvious that i would need:

Ten small double cardan axles, all extendable (they must be able to do so, because the distance between the axles vary when the suspension operates).
Six servos for the steered axles.
One servo reverser, as steering on axle seven and eight are reversed.
Two gear boxes (or one gear box and one reducer, gear ratios will be calculated later!).
One motor (type and power will be decided later).

  Finding the minimum and maximum length for each was quite tricky. It is important to remember that minimum length cannot be compromised, while maximum length will have to looked upon with forgiving eyes. Limiting axle / suspension travel will reduce the maximum length needed.
Then, finding the cardans to accomodate that is the next challange, but I got it done in due time.

And here they are! Not cheap, but cool! (three of them stayed in their sleeping bags as they are oiled up pretty good.)

  A small close-up of the cardan between axle one and two, showing how this extend-feature works.

Everything is steel, so they should be up for the job. Time will tell...

  The steel core was stripped of everything else, and flipped over. Axles mounted, and fitting could begin!
By mixing all 10 x 2 halves around for a while, I ended up *NOT* having to shorten any of the cardans at all. Excellent! A few axles have to have their up/down travel limited so the cardans do not pop out, but that's a job for later when the drive train is complete.
First I must make two housings for the bearings for an axle that will extend over the front most outriggers, between axle three and four, and build all steering mechanism.

  Tadaa!! My CNC machine! Making parts are now in another century:

1) Draw the piece in "Inventor" (software), export as DXF.
2) Open that file in "Cut2D" (software) and define tool paths etc, save work as G-code.
3) Open G-code in "Mach3" (software), turn on CNC, and run.

Well.. simplified! There are 1.000 things to setup and configure, possibilities are endless.

  This is my first pieces that I ran on the CNC: Steering tracking rods for the axles, CNC cut in 2mm brass! Note the handmade prototype on top...
I could never make components as accurate as the CNC.. nor as fast! These eight pieces took less than 10 minutes to cut, once developed, drawn and setup.. and they are accurate within 1/100mm.

Two differ from the rest because axle five and six are to be fixed, while all other axles are steered.
I will be placing one servo on each steered axle, six in total.
(For that I made six brackets in 2mm aluminum, pretty straight forward, thus no images)

  And done! All axles are now either locked, or equipped with individual, 10kg servos.

  This is the front boogie, and axles one, two and three.
All steered.

  This is the middle boogie, and axles four, five and six.
Axle four is steered, axle five and six are locked.

  This is the rear boogie, and axles seven and eight.
Both steered.

  The core seen from behind, towards axle eight:
The steering servo and bracket fits perfect inside the steel core, and even help keep the axle in line!

The pendulum suspension does allow some sideways travel, but the servo bracket limits this sideways travel to a minimum. Perfect..
-and to honest: I did not even plan for this double role of the servo bracket.. I came for free.

  Axle eight at maximum down position. Servo bracket still help guide the axle, and limit sideways movement!

  Axle eight at maximum up position. Servo bracket, and servo, fits nice an easy up against the steel core.

Now the cardans and steering components are in place! Let's line it all up!

  This is how I track the steering:
A pink string is tied to a point right between the two non-steered axles, as this is the point around which the vehicle will turn.
The other end is held over the front most, fully turned wheel, while observing the angle between tire and string..
The point in the lower left hand corner is the center of the turning circle for the vehicle, and this point is moved back and forth until the angle mentioned above is as close to 90* as possible.

  A close up of the angle between the tire and the string.
When this is completed for the front most wheel (that will turn the most), then it's time to look at the other axles, WITH OUT moving the center!!
Instead, steering linkage is adjusted for every axle so that a) wheels are aligned when servo is in center position, and b) the string - wheel angle is 90* when fully turned. This ensures that all wheels steer around the same turning center, thus reducing drag, resistance and power consumption.
(A trail run on a dusty surface should leave perfect tire tracks, not smeered, if this adjustment is ok.)


Connecting the cardan joints was pretty straight forward... until I had to run the drive train under the front outriggers that goes between axle three and four! The outriggers will completely fill the big open hole in the chassis, so that space is totally reserved!

I had this aluminum block made, and fitted a shaft, two ball bearings, and two cardan joints.

  The bearing house then went on the crane, shown here up side down.
The house is bolted on to the steel structure that goes under the outriggers ("support legs").

  Shown right side up, and the bearing house in it's final place. It can withstand the twisting motion from the working drive train, as it travels onwards to axles one -> three.

(Axle four is the one on the left, axle three is the one on the right.)

  With most of the cardan in place, I focused on motor and gearboxes.
This is the initial test to see how much "pull" is needed to move the crane.
The chassis was loaded up the 41,6kg, which is my estimate of the weight of the finished crane.

Pulling that on a flat surface required 2kg of force.

  The track where I will run the crane has climbs up to 10*, so I made a 10* ramp and repeated the pull-test, again with full weight.

Pulling up hill on a 10* slope required 10kg of force.

  A close up of the method.
With my finding I calculated the necessary gearbox output (torque) to the cardan axle, and came up with 200Ncm.
From that I could find the minimum motor power on the other side of all the gearing, and in turn the speed as well.

How? Let me show you!

  I have made a spreadsheet in Excel, but the procedure is pretty straight forward:
First the power:
The force needed to move the crane up a 10* slope is 10kg.
I then calculated as if the crane had only one wheel, so 10kg * wheel radius 5,4cm = 54kgcm of wheel torque. I then divided that by the gearing in the differential: 54 kgcm / 2,67 = 20,2 kgcm. That, converted to Newton, is close to 200 Ncm of necessary torque on the cardan shaft to climb a 10* slope. From there it's only a matter of calculating your way through the reduction box and the gearbox, to the motor: 200 Ncm * 4 * 22,65 (1. gear) * loss factor 1,25 = 2,7 Ncm (The absolute minimum of torque that the motor must be capable off.)

Then the speed:
First the minimum max speed of 0,29 m/sec was converted to 1770 cm/min. 1770 cm/min divided tire circumference 33,9cm, multiplied by the gearing in the diff' 2,67 = 139 rpm on the cardan. Multiplying that by the reduction gearing 4, multiplied by 3rd gear ratio 5,66 = 3154 rpm.
(Loss factor does not apply to rpm) (The absolute minimum rpm that the motor must be capable off.)

However: Choosing minimum values would cause a lot of gear change to get either the power, or the speed needed. I therefore went for a motor to fit 2. gear specs. I found an motor from LRP called Truck Puller 3, 12V version: rpm is 6300, 36,1 Ncm.
Plenty of power, and speed high enough to give even higher speed, without turning the crane into a Ferrari. I won't go brushless because most of the time the crane will sit at a site working, not drive around. Efficiency is not important, and I REALLY like the Servonaut M20+'s auto-throttle!

  With the new motor data, I went ahead and calculated the other way around: From motor to road!

Basic data was the same, and calculating was more or less a question of multiplying or dividing the gearing, and looking at the wheel radius and circumference. The result is that I will have about three times the necessary power, and a top speed of 2,12 km/h in 3. gear.
In reality it looks like I will be using 2nd gear most, with the option of raw power in 1st gear, or turbo speed in 3rd! ----> Excellent!!!!

NOTE: These two calculations are available as online versions here: Vehicle Engineering Tools

  I solved an unforeseen problem while waiting for funds to get the motor + gearbox:

The extendable cardan joints will slip out of reach if the axles drop too far from their neighbors, so limiting the travel of the pendulum suspension is in some places critical. The chains serve that purpose where needed. With that problem solved, I can begin to fit motor and gear boxes.
I got the idea from an old Volvo. The rear axle drop was limited with straps. Good enough for Volvo, good enough for me. :-)

  Tadaa: Truck puller #3, 12V edition and Veroma 3 speed gearbox. That's 6300 rpm, 36 Ncm into gears: 22,65:1 or 11,32:1 or 5,66:1, and then into the 4:1. Lots of torque, and more than enough speed.


  The motor and gearbox needs to go into the 4mm steel core of the chassis right above axles five and six, as these two axles does not have steering servos.

First four 10mm holes marked the corners. To add strength I want the corners to be round, and not sharp.


The cut was done using my electric jigsaw with a steel cutting blade, oil, and then filed by hand.

  Then a bracket was made for the motor assembly, so that the motor and gearbox sits half way down through the chassis.
A servo for the three gears will be fitted later, after the slewing bearing has been fitted.

  Seen from below.

  Again, seen from below, - and with axles four, five and six fitted to show how it works.

The suspension is not limited by the motor and gearbox, and the shaft exiting from the gearbox will travel to the space right under the slewing bearing, where the 4:1 converter will be fitted, and connected to the axle drive train.

  Next: I need to cut the plastic deck of the crane, so the bracket with the motor and gearbox will fit inside the center column of the cranes back.

It will be made possible to take the motor and gearbox out this way for service and lube.
(Gearbox must be lubed once and again...)

  A little paint, small improvements, and cutting the plastic deck later..
Motor + gearbox fitted in the "engine compartment".

  The grill was milled on my CNC, and allows cooling of the motor.
I know that the real crane does not have this grill, but I have to deviate from that from time to time, as this is a working model.

  Next phase is the 4:1 gearbox / torque converter, reducing speed by four, and multiplying torque by four.
I made the sketch in Inventor, and applied the CNC tool paths in Cut2D, and made this screenshot.

The gear box house will house three 6,5 mm wide steel gears, six bearings, axle holes in three places, and two plugs for chancing / adding oil. (Shown later)

  Here's the late stage in planning, and gear selection.
I've been collecting gear for a while, and will need a lot of them for later components on the crane.

  I have made 2D parts on my CNC lots of times, but I have never made a 3D part on any CNC before, and never milled aluminum. There is a first time for everything, so let's go!
The 12mm aluminum sheet was bolted to the CNC machine, and I was close to hitting "Go!"... when I discovered that the CNC control software "Mach3" has a limit of 500 lines of G-code when running in demo-mode.. I have divided the entire process into three parts, but still the biggest block is well over 5000 lines. The project will continue when I receive the license.

  Several days later:
I got the license, AND broke the CNC program into 28 separate jobs, and the first is shown here.
I opted to make the parts using many small jobs, rather than one huge one, so it would be easier to recover from eg. tool breakage or so.

  After the first day I had this. The CNC was set to 10.000 rpm, 4mm end mill with two flutes, and the feed was 1,6mm / sec. Z axis was set to 0,6mm pr. pass. I used alcohol for cooling (of the CNC, not me... I had coffee!), and had frequent brakes to clean up, ventilate the room, and make sure that I was still on track.

  Day two! First little steps..

  End of day two! Here it's easy to see where the six bearings will go, and the index libs are also shown.
Little holes indicate where I need to drill and cut M3 thread later.

Final step on the CNC is to cut the two parts from the 12mm aluminum sheet, but I ran out of time for now.

  Done with the gear housing!
I am pretty happy with the result of this first time ever 3D CNC job of mine. :-)

  The assembly seen from the front. "In" is the top hole, and "out" to axles 1 -> 6 is the lower hole.

  The assembly seen from behind. "Out" to axles 7 and 8, and the two threaded (M4) holes for oil change / check.

  Here (a while later): All 4:1 torque converter components!
Gears and axles are solid steel, all fitted using six ball bearings.

  Ready to go on the crane.

Seen from the front, "in" at the top, and "out" to axles 6, 5, 4, 3, 2 and 1 below.

  Seen from the back, "out" to axles 7 and 8 below.

  Fitted between axle 6 (to the right) and 7 (to the left)

Not shown in the picture is the shaft from the torque converter and to the 3 speed Veroma gearbox. It will be fitted using a couple of flexible joints, no big deal.


The final detail on the drivetrain etc: Gear servo.
The servo will go into a hole in front of the forward outriggers using a steel servo bracket....

  ... and a carbon rod will connect it to the 3 speed Veroma gearbox.

That completes cardans, steering, gearboxes, main motor!
Next is outriggers and slewing bearing... and various details here and there.

Back to index --->