Aidan Williamson Log

From Open Source Ecology
(Redirected from Aidan Log)
Jump to: navigation, search

contribs
If a link to CAD takes you to DropBox:

  1. Open the link
  2. Top right of screen, download directory or download folder
  3. Unzip the .zip file
  4. Open main assembly file with FreeCad
  5. With A2+ workbench you can edit sub assemblies or individual parts by clicking on them in the main assembly window.

Old Logs:

October 10, 2019

Went to Ram's farm to scout dirt. We got three samples. They were all wet as it has been raining a lot here. Two were in piles at the kind of main farm area at the property. They had been dug nearby. One appears to be a montmorillonite clay because of the amount it swelled up when doing a jar test. It is similar to the lagoon sample from near copper bank. There was some large, coarse sand and larger crystal-like rocks distributed throughout it. The second sample was similar in color and showed some plasticity but seems to be similar to marl as it it much less plastic than the clay sample. The third sample is from a ways out in the farm and is sandy with (hopefully) some clay. When damp it can be formed into a ball that does not fall apart without a poke. This may be a sandy-clayey soil suitable for cement stabilization. The samples are being analyzed.

October 9, 2019

Updated the File:CEBPressJuneGroup.fcstd file to reflect the 16.09 machine. Wiki won't let me upload the 4mb file so here's a dropbox link: CEB16.09Belize.FCStd GitHub Link

Changes from File:CEBPressJuneGroup.fcstd:

  • Drawer guide system added and old version removed
    • 1" drawer guide rods added
    • 1/2"x3"x18" flats added to drawer
    • Drilled nylon bushings added
    • 1" nuts added
  • Main Cylinder extended to 16.5"
  • Main Cylinder mount plate added to frame
  • Shaker system deleted

October 8, 2019

I had the stroke limiter made at the local Mennonite machine shop. They didn't have any solid piece of 5" diameter material so I had to go check the nearby scrap yard. There I found a piece of 4.5"OD, 0.375"wall tube at about 10" length. The guys at the shop machined some 4" round to be a press fit inside the tube which was also machined. It took them about 5 hours to make the part while they also did various other jobs that other customers brought in. The charge was $100BZ, $50USD. I paid $10USD for the piece of tube, less than a dollar USD for the 5/16"x2" grade 5 hex head bolts and $3.00USD for the o-rings.

October 7, 2019

Tried doing some WebGL stuff yesterday. I can't seem to get any of the loaders to work. No idea what i'm doing in the javascript department but editing the HTML is easy enough. I am pulling examples from the threejs.org website but they don't work with my models. The benefit of doing a loader - be it .OBJ or .glTF - is that your file sizes are smaller. This is important for someone like me on a data connection. For now the FreeCAD WebGL export gives the best looking files though they are larger.

The stroke limiter will have to be 5.5" in height. They have 2.5"ID o-rings at the local hardware store. Hopefully the machine shop can make this. The only tricky part for them will be the counterbore for the socket-head cap screws. I belive they will have a chunk of steel the right size.

I made a drawing for it with TechDraw workbench on. Some features are missing but it's better than the Drawing Dimensioning Workbench that I had been using. Apparently it is being improved upon right now and the next FreeCAD release will have better features. Read about it here: TechDraw FreeCAD Wiki

I don't have a printer so thats why I'm doing the WebGL stuff. It's useful to show the guys in the machine shop since they don't have computers.

Stroke Limiter 5.PNG

The way I would make this:

  1. In the lathe, drill and bore the 2.5" clearance hole
  2. Cut the clearance chamfers on either end.
  3. Cut the o-ring grooves. These are not critical
  4. In a mill, machine a flat where the the counterbore will go for the 4 holes
  5. Center drill on that flat and then drill out with a #7 drill.
    1. Optional: now you can drill out halfway and counterbore the hole so that you can start your tap on the mill if you have that capability. This will make the tapped hole line up perfectly with the clearance hole.
  6. Part down the middle with a bandsaw
  7. Tap the tapped half if you haven't already
  8. Drill out the holes and counterbore the other piece if you haven't already.
  9. Make some shims depending on the thickness of your bandsaw blade.

October 6, 2019

The reason we can't just use a shim to shim up the press foot 2" is because then we would have to move the press cylinder down due to the overhang at the top end of the stroke. We have to limit the down stroke so that only 8" of material falls into the chamber. .

The drawing below shows a height of 2" this will actually be more because there is some rod exposed (painted black) before the stroke limiter engages the cross welded tube.

October 5, 2019

Fabricated a-frame type soil screen. It is immediately very useful.

First trial run of stabilized blocks ended in failure. It took me over a month to realize it but the machine has too much stroke on the main cylinder. It is a 10" stroke cylinder instead of an 8" stroke cylinder that the rest of the machine was designed for. So the blocks emerge between 5.5" and 6" in height and undergo an unconfined compression cycle as they are ejected. This causes them to crack and deform immediately. The drawer then destroys the blocks as it retracts. Terrible.

I need to stroke limit the cylinder. The way we used to do this with pneumatic cylinders was to put a machined disk inside the cylinder to set the stroke to whatever we wanted. I want to limit the retraction so I'm going to use an external stroke limiter that will sit on the rod above the cap end of the cylinder. It should have o-rings or be made of a soft metal as to not damage the rod. Here is a drawing of what it would look like if I were to make it in my machine shop. This would be machined from 5in 6061 round stock. Obviously there are other ways to do it. It could be a square profile. The bottom line is it has to sit on the rod above the cylinder, not damage the rod, be able to be installed without removing the rod, and be strong enough to withstand the forces.

The test blocks were made with 20% Sand, 10% Cement, and 70% dirt. The water content was decided qualitatively with the "poke test". I think lower soil:sand ratio is desirable.

September 25 - October 4, 2019

Moving block press site to new location. Setting up site. Fixing my and Scott's trucks. A cane grabber came to unload the block press from the trailer. New site doesn't have electricity yet but it does have running water.

September 23, 2019

Playing around with WebGL stuff. Here's a VERY primitive workflow:

  1. Starting with A2+ assembly file
  2. Select all on object tree
  3. Export to WebGL.
  4. Put resulting .html file in a folder with a subfolder containing three.min.js
  5. Open resulting .html file in a text editor
  6. Edit the line where it gets the three.min.js source file as per Michel's instructions.
    1. Make sure you use three.min.js Revision 50 if you want to go directly from freecad to webgl I upgraded the version on my github.io repo and it caused an issue with trackball controls. I think orbit controls may be superior but the freecad exporter uses trackball
    2. Also can add antialiasing if desired. See video:
    3. default.jpg
  7. Optional: make an index.html file
  8. Create a github.io page
    1. In your github account: Make a new repo with name username.github.io
      1. My username is CabbageBreath
    2. Upload your project files to that repo
    3. https://username.github.io/modelname.html is where the project will be hosted
    4. My index is at [1]

MJ Sez - Nice, results from basic export are quite decent and load pretty fast.

September 22, 2019

Modeled the pendulum crusher. I bought a truck wheel rim to cut in half to make this little beast. The CAD is available here: Pendulum Clay Crusher DropBox

The fellow at Parry Building Products uploaded a video of their design which is based on the Intermediate Technology Workshops (ITW) model:

default.jpg

I also modeled the Modulus of Rupture (MoR) test jig specified in the Compressed Earth Block Manual of Production. This uses CEBs placed on a platen to apply a force for a three-point-bend test to measure flexural strength of blocks. You can relate this to compressive strength by multiplying by a factor of between 3.5 and 8. The Germans give 5-8 as the factor whereas Krosnowski found it was between 3 and 4.

File:Block Breaker.FCStd

Block Breaker CAD DropBox

September 20, 2019

Playing around with sensors on the brick press. The hydraulic system doesn't respond as quickly as I'm used to with pneumatics. This may be due to the landing geometry on the solenoid valve spool being that it is a motor valve. A mid-position sensor is useful for my idea to have a button on the controller case that simply brings the drawer to the mid position no matter where it started from.. I'm using IFM Effector m12 style inductive sensors that I found used in a bundle of 8 for $60 on eBay. Cables I got from automation direct. They're pnp 12v. Works fine with the arduino.

September 19, 2019

Welding repair of Scott's riding mower deck

September 18, 2019

Sieve testing LB soil. CEB_Microhouse_Build_in_Belize/Notes_9-18

September 17, 2019

pH tests for lime stabilization. A 10% stabilization gave 12.1pH. The Little Belize clay may not be suitable for lime stabilization based on this test result.

Did the 10% test twice

Lime Percentage Weight of Soil (g) Weight of Lime (g) pH
2.5% 24.375 0.625 11.7
5% 23.75 1.25 11.8
7.5% 23.125 1.875 11.9
10% 22.5 2.5 12.0/12.1

Used the cement mixer to mix LB soil with sand. It works well for smaller batches. Good for mixing the test blocks.

Started welding up the soil screen which is just an a-frame with expanded metal that you screen the dirt through before mixing.

September 16, 2019

Running around Corozal getting immigration paperwork done for another month. Having my truck here makes it more diffficult.

September 15,2019

Integrated automode into manual button code. I'm taking a similar approach to 2014 to avoid relying on while loops and delays. File:Controller v19.01 Belize.zip

Updated logs. I have poor access to internet here so I just write notes on my smartphone. will try to get better about documentation and uploading multimedia.

September 14,2019

Went to L.B. to talk to Bernhart about setting up the excavation of his site. Aiming for 24th. Looked for dump trailer options. Painted CEB press with Scott, Cecilia, and Evelyn. Finished sieve collection box. In LB went to an auto salvage yard where they have many heavy duty truck parts for sale. I will to to make use of this "graveyard" to source parts in the future. Bought a wheel rim for $15 for the pendulum crusher prototype.


September 13, 2019

Still getting the high pressure issue. Pressure builds, the powercube bogs down, the gauge drops to 0, and the powercube stays loaded and the cylinders don't move. Thinking I got some dirt in the system and the valves are stuck, I take them off the block and look for debris. I also check that they are shuttling when powered. They look great.


Afraid it was due to the pilot operated check valves I installed so I removed them and reinstalled system to original spec. Same issue.

Took the fittings on the manifold input off and I found that the input quick-connect fitting (female) had failed and that pieces of metal were in the system. I found all the broken pieces (i think) and replaced the quick-couplers with a 1/2"npt coupler.

The system works great. Reinstalled pilot operated check valves and the system works greater. The cylinders have good speed and they don't drift.

I could have installed the EATON VICKERS solenoid valves we have here but they draw more current than the relays on the HI LET GO relay shield are rated for. A lower current draw cylinder valve would be more expensive than the pilot check valve system. Also the motor valves still give more flow.

September 12, 2019

Fully integrated the pilot-operated check valves. They are now installed on the manifold. This required me to flip the ports so that the retracts are on the A ports and the extends are on the b ports. Just a piping issue. I simply switched the wires in the solenoids.

Worked great until suddenly it didn't. I would get high pressure and then not be able to change direction. I thought the solenoids were getting stuck as if I accidently let some dirt in to the system when I was doing the modifications or there was an electrical problem. I opened the covers and saw there was water damage in the electrical connection. Upon taking it apart the screwheads stripped since the insert nuts they interface with were rusted. The heads were super soft and stripped very easily. Only one solenoid connector was damaged. Lots of solder later it works great. Lost a flyback diode in the process. Running without one right now.


September 11, 2019

Trip to Orange Walk. Acquired 8kw gas generator. Caulked the lowermost sieves.

September 10, 2019

Installed pilot operated check valve on the press cylinder. This was installed temporarily on the hoses that lead to the press cylinder. Works great. 1/4" signal hose had a pinhole leak out of the box from surplus center. Don't seem to reduce flow too much.

Pressed a test block with the little belize clay. Cut it with some sand. Still setting up block testing so nothing scientific yet.

September 9, 2019

Manual control buttons fully integrated. Automode switch installed.

September 8, 2019

For the buttons I initially wanted to switch 12v since the 5v regulator had already failed once. I'm now using an LM7805 regulator with +12v Positive and Negative leads to the bus bar and the 5v output put into the +5v rail on the double row header. Looking back it was a bit more than necessary but I wanted to switch the 12v with the switches and step it down with a voltage divider directly to the output pin for the relay. The issue is that when a pin is configured as an output it's input impedance changes and messes up the voltage divider. I didn't want to go through the microcontroller for the buttons but I couldn't find the right amp-rated momentary switches in time so I had to order the ones I got. They're IP67 rated momentary switches that are NO. They switch GND to inputs configured with internal pullups.

I had an issue with a placement of a parenthesis in my program that cost me a good hour (i'm not proud to say). Changed up the controller box by switching the outlet box used as a spacer with a piece of wood so that I could fit more arduino shields on the arduino.

I bought a shield for doing IO prototyping. This is the one I got: link. I don't recommend it. It's fine but I would get a simple protoboard without all the bells and whistles. Something like this I also got these screw terminals link and some 10k and 18k resistors for 12v to 5v signal conditioning.


September 7, 2019

Went kayaking.

Started adding controller buttons for manual control.

September 6, 2019

Shopping. Went to Ram's place to get parts from USA. Electronics and hydraulics.

September 5, 2019

Picked up Scott

September 4, 2019

Finished sieve boxes. Adapted vibrating pedestal thingy to work for vibrating the sieve boxes.

September 3, 2019

Went to Bernhart's house. Bernhart is a Mennonite fellow that works at the LB hardware. He invited us over because he wants to dig a pond and will give his clayey subsoil to us. He wants us to pay for the excavator at $125/hr. It's 18mi to Copper Bank from the dig site.

September 2, 2019

Finished painting cement mixer. Worked on sieves.

September 1, 2019

Made more cuts for the soil sieves. I feel strongly they should be made with notches so I'm waiting to have Noae cut the notches on his tablesaw. Once the sieves are done the work on the soil testing can begin.

Prepped and painted cement mixer with Harrol. Ran out of paint 3/4 the way through.

I crushed the lagoon clay and ran a jar test with it. I had to stir the mixture to get it to mix properly. Shaking did not suffice. The next day the specimen had swelled up to twice the marked level. Extremely expansive. I need to test this more thoroughly.


August 31, 2019

Day off. Did some A2+ style modeling of Soil Mixer 2019. Need to do hammers and then start on mixing chamber.

August 28, 2019

Trip to Corozal. Got paint for cement mixer. Tablesaw motor brush wore out. Glue repair held strong. New brush on order.

Went to Noe's house to rip the sapodilla wood for the sieves. He did a fine job and only asked to be compensated a dollar. He built his own tablesaw.


August 27. 2019

Got the machine to run pretty well with some code adjustments. I now drop the main cylinder for a half second before moving the drawer to the eject position. The machine would give high pressure if I didn't do that and the ejection would fail. Haven't tested it with soil in yet.

Repaired a melted tablesaw motor with some JB weld so that I can make the soil sieve stack.

Went looking for more clayey soil. It's mostly marl in the immediate vicinity. I ended up bringing home a wheelbarrow full of some clayey topsoil. Not ideal but better than nothing. I need to go to Little Belize and figure out where their clayey road soil comes from.

The marl test block we made seems to be getting stronger as it cures. It deserves some closer attention.

August 26, 2019

Worked on aligning the brick press and tightening any fasteners that may have been loose. I played around with putting shims in to get the drawer to be square with the machine before realizing that it really doesn't matter since the squareness of the bricks depends on the inside of the press chamber and not the drawer. The drawer just provides the top surface. THe only thing that matters with the drawer is that it doesn't deflect when pushed up against so that the top and bottom faces of the block are parallel.

Later in the day we pressed some blocks using the marl beneath our feet. Screened with 1/4" screen and varying amounts of moisture added. No objective measurements. Just a little ball test to get in the right range. We pressed the blocks manually since the auto code still isn't optimized. This is a tedious process that requires turning on and off the power cube and disconnecting hydraulic hoses and using a jumper wire to turn on and off the solenoids. First extend the drawer and drop the main cylinder. When the main cylinder is at its limit turn off the power cube and disconnect the main cylinder hoses. Then turn on the power cube and move the drawer to the obstruct position. Then turn off the powercube, reconnect the main cylinder hoses, turn on the powercube, press the block, drop the main cylinder all the way and move the drawer to its home position. Wait for the block to rise and then push it out. Very tedious process and jumping the wires makes it even more so. We pressed three blocks. THe first one had enough clean surfaces to see how square the block was. It appeared to be square to a 16th of an inch on most sides. The marl doesn't have enough binder in it to work as a good CEB material. It shattered with a light punch. We added more moisture to the next blocks but they were too wet and water seeped around the block as it was ejected. Still it held its shape and we left it to dry.

At some point the 5v regulator stopped putting out 5v and the arduino won't drive the relays anymore. I have to have a usb connection to make it work.

August 24, 2019

Reversed logic of pressure switch in code to match reality. It's a NO pressure switch and we are using pullup resistors so we are feeding 5v on pin 13 to one side of the switch. The other side of the switch is GND so when we connect pin 13 to GND we should get a LOW on pin 13. The program was looking for a HIGH instead of a LOW.

We adjusted the pressure switch with an ohmmeter and a jumper cable. Since there is no provision for manual control we had to jumper 12v to a solenoid with a wire to trigger a high pressure for setting the pressure switch. Connect the ohmmeter to both sides of the switch with the leads disconnected and trigger high pressure. While creating high pressure in the system in short bursts, tighten the screw until it no longer reads connectivity and then back the screw off until it just reads connectivity at high pressure. It's a 3mm hex wrench in the pressure switch on the machine.

Added serial communications to the code for debugging. Each state of the cycle will generate text indicating where it is. As does high pressure.

First real issue was that if you wait a long time to switch to press mode from reset mode after the reset homing is complete then it will think that it took a long time to extend the drawer. It then scales this long time (15s or more maybe) to retract to mid-point but even 5 seconds of retraction will take you to the end of the stroke so you will bottom out and the code doesn't look for high pressure. It's just delaying. This is the issue with delays. Eventually I will rewrite the code to do away with delays and while loops but for now I just band-aided that step by replacing the delay with an if and while statement.

This is the original:

                                                 //Step 2 Compression Chamber Closure
if (resetSelected() == false) {
 digitalWrite(SOLENOID_RIGHT, HIGH);
 delay(drawerExtTime*0.75*0.5);           //0.5 is the half-way point of drawer cylinder; 0.75 is the Drawer Multiplier
 digitalWrite(SOLENOID_RIGHT, LOW);       //Can modify the 0.5 to fine tune the drawer closure location
}


This is the version that will move to mid point or until high pressure is read:

//**********************************Step 2 Compression Chamber Obstruction*********************************************
if (resetSelected() == false) {
 Serial.println("Moving Drawer to mid position");
 if(digitalRead(SOLENOID_RIGHT == LOW) ) {
   Serial.println("Current Time:");
   Serial.println(millis(), DEC);
   obstructMoveTime = drawerExtTime*.75*.5 ;     //Duration that the solenoid will be on 
   Serial.println("Obstruct Move TIme:");
   Serial.println(obstructMoveTime,DEC);
   drawerEndTime = millis() + obstructMoveTime ; // calculate the time the output should turn off
   Serial.println("drawerEndTime Calculated");
   Serial.println(drawerEndTime,DEC);
 }
}
 Serial.println("here1");
 while (millis() <= drawerEndTime){
    if(checkPressure() == true){
       drawerEndTime = 0;
     }
     digitalWrite(SOLENOID_RIGHT, HIGH);
     Serial.println("here2");
     Serial.println(checkPressure(),DEC);
 }    
 
//delay(drawerExtTime*0.75*0.5);           //0.5 is the half-way point of drawer cylinder; 0.75 is the Drawer Multiplier
digitalWrite(SOLENOID_RIGHT, LOW);       //Can modify the 0.5 to fine tune the drawer closure location
Serial.println("Drawer at mid position");
delay(3000);                             //Delay for slowing operation down for debugging

It's not beautiful but it works for now. In order to do the digitalRead on the solenoid output I had to change the #DEFINE SOLENOID_RIGHT to an int SOLENOID_RIGHT = 0; so that its state can be read.

I also dropped the 20ms pressure switch debounce to 5ms. This appears to help timing overall. I can reduce it even further. 5 was arbitrary. I think 2ms would do just fine. Gotta research that but i don't think contacts bounce for more than a ms.

The drawer extension time measurement is consistently too short. Common values are around 1.7s. This means that the chamber doesn't obstruct fully. This is due in part to the fact that the drawer cylinder isn't starting at its actual home position. High pressure on the main cylinder (second in the series manifold) causes the drawer cylinder to bump forwards a bit more than an inch by the time it goes to extend from fully retracted to fully extended. So I know that this is causing short extension times. There may also be a problem with the way the code is measuring. Changing that debounce value shifted the times longer so that definitely helped.

I need to get the hydraulics in order (press cylinder moving forward due to main cylinder pressure events) but first we need to square the machine up mechanically. It's not square whatsoever. Next step is to loosen all the fasteners, align everything, and retorque the fasteners.

The powercube is running well. The CEB press seems like it will shake itself apart every cycle (as always) and Cecilia went running the first time I ran a cycle. Some hydraulic fittings are leaking. A vise is on my wishlist. Adjustable wrenches will have to do for now. Hopefully i'll get internet access soon to upload multimedia.

The clayey dirt from the mennonite roads expanded a bit more than the marl in a jar test. Still it's more clayey so cutting it with sand might yield good material. Soil testing has been on the backburner because I'm trying to get the press to make a block before Scott leaves on Monday.

August 22 - 23, 2019

Worked on repairing Scott's old broken cement mixer. Frame is bent and keyway in shaft is worn. Rody, Harrol, and I are reinforcing the frame.

Built a workbench to work at near the machine. Started looking at the controller and code. Found a potential issue on line 102. The instruction is to turn off the drawer cylinder instead of the press cylinder. I also noticed (and this is just a stylistic thing) that often there are if else statements when a simple if statement will suffice - there are no arguments in the else portion. I'm going to have to update the code when I add the manual buttons so i'll upload my version when the time is right.

I think there is a need for manual mode because if someone gets their finger or arm stuck in the machine there is no way to move the drawer to help them.

Scouted soil at Scott's first 60acre jungle property. Clayey soil under a decent topsoil layer. Didn't take samples but will do so at a later date.

August 21, 2019

Went to "Little Belize" which is a Mennonite community where they have a lumber yard and a hardware store. The hardware store sells some metal. I bought 40' of 1/8" x 1 1/2" angle, some flattened expanded metal, and some 4x4 screen to make a screen for sieving soil for production. I also picked up some 3/4" x 4" x 84" hardwood from the lumber yard for making the sieves. I brought down 12"x12" soil screens from mcmaster to reduce the price of a soil sieve set. The lumberyard has SO MUCH HARDWOOD!!!! SO CHEAP!!! It's incredible. and they let you pick through the reject lumber for free.I got what I need to make the sieves and a workbench for $30 US. And this is hardwood, not some shitty pine. I'll upload some videos of their sawmill. They planed the wood for the sieves for me. I will start making them soon.

Spent the rest of the day testing power tools. Tablesaw motor is melted at the brush housings. I'll attempt to fix with epoxy. We have several working circular saws. Gabi Harrol and Scott set up some tarps for sun protection while working.

Adjusting to the heat and sun. It's not much different from Missouri in July but my access to water is different and I'm on other peoples' food schedules and I have an issue with a swollen tonsil that is exacerbated by the amount of talking I have to do. Can't complain though since we're in paradise.

While in Little Belize I grabbed a sample of the clayey soil that they use on their roads. It looks more promising than the Marl. Haven't jar tested it yet. I'll need to get a larger sample after I get the sieves going.

August 20, 2019

First week in Belize. Internet access for now is a hotspot on my phone. Not enough data to upload multimedia yet. There is wifi available somewhere in the village but I haven't been introduced to that person yet. We arrived in the afternoon on Saturday. Spent Sunday and Monday getting settled and taking care of the basic necessities of life. Brought all of Scott's tools to the worksite at Harrol's house.

We successfully unloaded and setup the brickpress and power cube. Arranged them in what I think will be a good workflow. The powercube needed adjustment on the choke cable. The people down here who had tested it had been putting gas down the air intake to start it. I adjusted the cable properly so that the choke comes on when it should and the engine starts without any other intervention. Oil was clean and level looked great. I believe Ram changed it. Hydraulic fluid tank was 3/4 full. Air filter was pristine. Won't need the spare parts for a long time.

We hooked up the hydraulics and powered on the brick press with the controller. It didn't seem to be doing what it should be doing. I need to check the setting of the pressure switch to make sure it's not triggering early. The machine seemed to do the RESET procedure properly and repeat every so often. But when I switched to full brick mode it didn't act right. It extended the drawer all the way and then jerked its way back in maybe 1 or 2 second increments. I didn't have my computer handy to check what the machine thought it was doing. I would have to put some serial communications in the code anyway. I don't have a good workspace set up yet so that's the priority right now.

One thing Scott wants to do is make the machine work in a manual mode. There is no way to do that right now besides using a jumper wire to give power to the solenoids inside the controller box. I don't see any manual override on the valves. The 2014 solenoids did have a manual override but I don't see it here. I ordered some momentary switches that should arrive next week. Later on I remembered that I should have gotten NC switches (best practice) but the NO ones I got should work.

I see that there are spare cylinder (tandem spool) valves down here. They're the Eaton VICKERS brand that Surplus Center sells. I have the hirschmann connectors on order. The issue with these is that they are 3.5 AMP coils. It says 30W on the housing but the catalog says 3.5A. I'll test what it really draws. The relays on the controller are rated for 3A. The relays might work at that current since they're not on for very long. We'll see.

Did a jar test on the marl that they build roads with. Marl is a limestone rich soil. It was pretty expansive. Came out to about 5/16" above my scribed line. Just a preliminary test. It may be useable.

August 6, 2019

I'm very interested in the pendulum crusher from the previous log entry. After a correspondence with the owner of Parry Building product, I'm convinced this is a rugged design and the output is claimed to be 20kg/min. If we assume 1800kg/m^3 1 then it would take 90mins to do a cubic meter or 70 mins per cubic yard. This gives 1.75 blocks per minute, roughly. Slow.

August 2, 2019

Clay crushers

  • Pendulum Crusher
  • ITW WORKSHOPS PENDULUM CRUSHER

  • See right side of image

  • Now manufactured by Parry Building Products.

  • Crankshaft crusher:
    • default.jpg

July 31, 2019

I ordered all the spare parts I think I may need in Belize. The orders are on page 2 of the Tools & Infrastructure spreadsheet at Mader's Machine. I didn't buy a spare pressure relief valve nor any spare hoses. Hopefully the relief valve doesn't kick on, doesn't get hot, and doesn't fail. I did find an alternate pressure switch, New Old Stock from eBay. Same logic and same port size.

Been thinking about the cylinder drift issue recorded at CEB_Controller_v19.01. I read that one way to correct the issue would be to use pilot-operated check valves on the press cylinder. These are $60 from surplus center and prevent a cylinder from moving unless powered.

I do see what Marcin is talking about with the "motor spools" being higher flow than the "cylinder spools". Yuken and Northman both give lower flow rates for tandem(C6) spools vs open center(C3) spools. For Yuken it's about half; for Northman it's about a third less.

I know now that motor spools are also called open center spools which is a use of open center that is distinct from the standard open center vs closed center usage. Yes, they're open center but so are tandem valves which connect P to T in center position but block (to the extent that metal seals on metal) flow from the cylinder to tank. Tandem valves should be used in combination with pilot-operated check valves in situations like loader arms so that cylinders only move when the valve is actuated. The pilot op valves would be a safety redundancy.

So I propose we add pilot-op check valves on the CEB press for a total of $155. This is less than an upgrade to higher-flow D05 valves and subplate in a tandem configuration. My quoted cost includes the additional hoses needed. I believe the check valves could be screwed right in to the subplate outputs. We would need to add Tees to port the pilot pressure to the valves.

July 18, 2019

Sketched up a rough outline of an auger-assisted soil crusher. Trying to solve the problem of relying on gravity to meter the dosage of soil and stabilizer. I think a force like that from an auger would help normalize the dosing. I'm not crazy about it but it could work. Maybe too expensive.


July 17, 2019

Don't have edit access on Soil Mixer Design Doc so I'll record my thoughts here:

  1. What propels the stabilized soil mixture out of the hammermill?
  2. Relying on gravity and ratio of areas to dose quantities of soil and earth. This, in my opinion, will be irregular. You would have to be able to change the size of the stabilizer aperture in the field.
    1. Based on pH for lime. Do a sample every so often
    2. Based on color for cement?
  3. The raw soil would need to be crushed and dried before it will fall into the drawer in a repeatable way. This kind of defeats the purpose of a hammermill.
    1. I could see this propagating the error we have in block size throughout the system. Relying on the soil to fall into a hole the same way each time. Might double the error, right? Double the random uncertainty in block size. Or give a systematic shifts in error during a shift or certain working condition.
  4. An auger system could meter the input of soil and stabilizer in a more controlled way while still using only one actuator (motor instead of cylinder - though maybe it works out to an extra motor). A variable pulley could be used to set the stabilizer ratio. This is an off-shelf item.
    1. Example of variable speed pulley system. If we use the same auger size for soil and stabilizer, then gear reduce the stabillizer auger to give 3% to 10% the speed of the soil auger. If you need more than 10% stabilizer then you are using a very low clay content soil and using a ton of cement - rethink the soil blend and if you need to, buy a different pulley.
    2. Does variable pulley work at low speeds?
      1. Planetary Geardown? 6 axis CNC Torch Table to produce most components of planetary geardown?

July 16, 2019

Tried to go to the dev meeting to discuss soil mixer per discussion with Marcin last week but nobody was there save an AFK Abe.

Need feedback on workshop tools and infrastructure list at Belize CEB. I'm not including any hardware or materials for the house - just the tools that are required to build it. Just working from what I can imagine would be necessary. Another more experienced set of eyes would be good to glance over what I have listed so far.

Thinking about soil prep: for Marcin's hammermill concept, how do we get dirt at OMC (optimum moisture content) before it goes into the hammermill? One way could be to figure out OMC by volume of dry dirt, volume of water and have laborers fill 5 gallon buckets with dry, screened dirt and add water from another bucket with demarcated OMC line. This unmixed bucket of water and dirt goes in to hammermill. Assume that the dirt has (non-uniformly) absorbed the water AND that the hammermill will actually give a homogeneous mix. I don't see why the hammermill would give a homogenous mix unless you cap the mixing chamber to prevent material escaping before a homogeneous mix is obtained. This would take longer than a press cycle so it would have to mix more than 1 block's worth per round.

July 11, 2019

Mader's Machine

July 5, 2019

CADded D3D Mill assembly in 2 days

June 20, 2019

June 16, 2019

File:Taper to Straight adapter.fcstd File:Taper to Straight Adapter Wheel Hub.fcstd

Feb 25, 2019

I made a new tube builder when I heard mine wasn't working in FreeCad 0.16 on Ubuntu.
The new file is based on the Sketcher, Draft, and Part workbenches whereas the original file was based on the Sketcher and Part Design workbenches.

File:Alternative Universal Tube Builder.fcstd

Here is a video explaining the new file approach

Feb 23, 2019

Video showing making of a spreadsheet parametrized cad file.

Feb 22, 2019

Made a video about the tube generator and improved the file by adding HoleSpacing parameter to spreadsheet.

Nov 20, 2018

Here is a version uploaded to the OSE Wiki. It should work but may have design flaws. I'm using an O-ring to seal the cylinder in the cap trepan groove. This is questionable.
File:Universal Cylinder Builder ADW.FCStd


I've been working on a cad file for the Open Source Cylinder. Here is a link to the alpha stage of the file. DROPBOX The end-caps are square with a trepan machined for the tube to seat. The caps could potentially be parted so that a welding operation replaces a machining operation for the tube seat. The Rod is welded to the Piston in my version.

Like the universal tube builder, I'm trying to make it parameterized from a spreadsheet. Here is what the parameters look like right now:

Name	Input (Inches)	Output (mm)	Note
Cap Side Length	4	50.8	
Cap Thickness	1	25.4	
Tube OD	3	38.1	
Tube Wall Thickness	0.25	6.35	
Piston Thickness	1	25.4	
Piston Seal Inner, Height	0.3125	7.9375	
Piston Seal Inner, ID	2.125	26.9875	
Piston Seal Inner, OD	2.5	31.75	
Piston Seal Outer, Height	0.25	6.35	
Piston Seal Outer, ID	2.125	26.9875	
Piston Seal Outer, OD	2.5	31.75	
Stroke	12	304.8	
Rod OD	1.5	19.05	
Rod Stickout	2.5	63.5	
Cap Seal Outer, Height	0.25	6.35	
Cap Seal Outer, OD	2	25.4	
Cap Seal Outer, ID	1.495	18.9865	
Cap Seal Inner, Height	0.25	6.35	
Cap Seal Inner, OD	2	25.4	
Cap Seal Inner, ID	1.495	18.9865	

Need to add a few parameters like tie-rod diameter. I think the model may need to be scrapped. Especially if one looks at the dependency graph lol. I didn't take as intentional an approach to modeling it as I could have.

I believe it is a worthwhile effort, though. A user could generate the CAD for a cylinder of different sizes by simply entering data into a spreadsheet.



Left

Notes

The caps as I have drawn can be easily made on a lathe and mill. How i would make this: Cut the plate to size and machine square. Surface grind for extra credit. Put the four tie rod holes in but make them half inch reams. Using shoulder bolts, attach the plate to a lathe jig or 2 bolts and 2 dowel pins. To make the lathe jig, press fit a piece of 4" or similar crs into a thick plate with the bolt pattern on it and weld. Face the jig on the lathe after welding. The bolt pattern should be put in after the welding operation. Use custom vise jaws and maybe a small bore for picking up the center. For the rod end cap, drill and bore the rod hole then bore internal grooves. Then cut trepan groove. For the base, it's just a trepan. Make those in half a day including making the jig.

As an alternative, I was imagining "slicing" the caps into layers, like 3d printing, and then welding the slices together.

For plastic cylinders, I would start with pvc or abs pipe as the tube. I'd go down that road before I started messing with a metal prototype.

I'll keep thinking about it.


Aidan

Nov 5, 2018

Here is a file that might be helpful for y'all. Open the Universal Tube Builder file and use the spreadsheet to create tubing with all dimensions parametrized. Holes will automatically be set to 4" spacing. I can change that if you want. You can define the standoff for the first hole.

File:Universal Tube Builder ADW.FCstd

FreeCAD expressions for constraint solver