Sunday, 13 February 2011

A momentus weekend.

I have spent the weekend working on the Mantis with Josh, who kindly gave up his weekend to help out with some of the more difficult tasks I still have to complete. First off we stripped down the test rig, and removed all the components we required to build the last leg. This didn't take too long and I can now empty out the rams and take them to be painted. Josh then started work on fitting the new load pin sensors to the ankles, and re-positioning the metatarsus encoders so that they can be adjusted with ease.

I set about fitting the 48 hydraulic valves and 12 pressure sensors to the manifold block. To do this we first had to drain the system down, and remove the hydraulic tank so that I had good access to the middle manifold blocks. The tank also needs work carried out, so I will do this now it is off the machine. It took a fair amount of time to finish but the valves went on well, and the chassis is looking really busy with all the electrical connections in place!

We then turned our attention to getting a set of rams back onto the right anterior leg. We did manage to get the metatarsus, tibia and femur rams in place, but by then it was 6pm and we were both shattered. Left the workshop looking a real mess and loads to do, but right now looking forward to a lay in tomorrow.

Friday, 11 February 2011

Hose assemblies

This afternoon I picked up the last of the hose assemblies from Unimaster. Its going to be fun fitting this lot!

Foot impact tests

The following videos were taken at 250fps to see how much bounce I was getting during the initial contact phase of the foot with various configurations of the foots inner foam support. The light levels were a bit too low for the camera but we hadn't realised until after we shot the footage!


In this first shot the foot has the foam support fitted. The drag at the end of the shot was me pressing the wrong key!! The foot loads up to approximately 400Kg.

Thursday, 10 February 2011

CanTainer finishing touches

One issue I had yet to address with the HexEngine enclosure (CanTainer) was the issue of IDC connectors vibrating loose and becoming disconnected. While looking at the PC-104 design this afternoon I decided to machine a couple of retaining plates and with the help of a slice of foam held in place with super glue the connectors are now secured from falling out!

Foam Inserts cutting and test.

Well the foam cutting for the foot inserts went very well today! The foam cut much better than expected on the CNC machine, although I did have to keep an eye on the process as occasionally a loose piece of foam would wind itself round the cutter which could cause issues.

I cut the foam in two parts, the lower part is more complicated, requiring several pockets to fit the complex shape of the rubber foot. The upper part was initially just a 280mm diameter circle, however I then decided to add the patter seen in the image. This is to allow air to escape through the breather hole in the aluminium top part of the foot, otherwise the foam blocks the hole when in compression.

I ran foot contact tests on the leg with and without foam inserts with a load of 300kg, the results were very good. The addition of the foam inserts naturally reduces the compression of the foot as desired. However, there is a little more bounce on initial contact. This may be aided with the air channels that were later added but have not been tested, alternatively a softer foam for the lower portion of the foot may give less bounce yet retain rigidity under load.

Wednesday, 9 February 2011

New HexEngine Enclosure

I have cut the new end plates for the HexEngine PC-104 case. This is a CanTainer which can take up to 5 PC-104 modules. The HexEngine currently consists of three:

1) PSU ( 12 - 24V in 5V out )
2) CAN Card (dual channel)
3) Vortex86 DX SBC

I have machined one end of the enclosure for 5 leds which indicate power and operational states for the software.

The other end has one 24v power led, reset switch and the following connectors which are all IP65 rated or higher:

1) CAN bus 1 + Power In (System CAN bus)
2) CAN bus 1 + Power Out (System CAN bus)
3) CAN bus 2 + Power In (Spare CAN bus)
4) CAN bus 2 + Power Out (Spare CAN bus)
5) LAN
6) COM1 RS232 (Linux Terminal)
7) COM2 RS232/RS485 (Will have RS232 Inclinometer)

Feet Foam

The sheet of EVA foam arrived this morning so I can finally cut out the inner filling to the rubber feet and see if this theory works! I tested the foam samples for compression, and worked out what I think the maximum force one foot could apply in PSI. I selected a foam that gave about a 50% compression at the maximum force. Not rocket science but with any luck it will help. I will post my findings tomorrow..

Friday, 4 February 2011

15Kw Heater

Its been too cold most days to get into the main workshop and do any work, so I took the plunge and purchased a 15kw 3 phase heater. I know its not going to heat this place up, but it takes the edge off of the area I'm working in. I worked out we would need about a 40kw heater for this unit, but with no insulation it would probably need to be on continuously! This 15kw unit was reasonably priced at £175 and was about the biggest I can plug into our 32amp 3 phase socket.

Thursday, 3 February 2011

Electronics / Loom installation

I have spent a bit more time working on the installation of each legs various looms and control box's VIC-E and SID. I have been putting this job of for a whil, partially because its been too cold to work in the main workshop, and partially because its a bitch of a job! Anyway, all the encoders are now wired up and connected and the looms are in place. All apart from the 5 leg which requires parts from the test rig once it is stripped down.

Installing the leg looms and wiring up the encoders.

Tuesday, 1 February 2011

A new team member for 2011

Despite my lack of posts since the New Year, (I cant believe its already February!) I have been busy working on the hexapod! I have now enlisted some help form a fiend / colleague Sandy White.

I have set Sandy to the task of coding the Operator Interface which will run on a GrayHill 3D65 platform running WinCE. While Sandy is wrangling with the code, I have been getting on with designing the graphics and screen layout for the OI. As well as more leg tests and improvements to the HexEngine.

I have also decided it would be really good to aim to have the Mantis take its first steps by the end of March. This should be achievable even if by remote control, as I don't think the OI will be ready by then. I will need to get some additional help to finish off the hydraulics install, hopefully I will get Josh back for a weekend here and there along with anyone else I can convince to muck in!