How to Build an Archimedes Screw

We are Team Shark, a group of Product Design Engineering students and as part of our uni course we've been given the challenge to design and build a water pump.  The task is to design and manufacture a pump that could lift 5 litres of water up a height of 600mm in a time of 5 minutes or less.  We've decided to make an Archimedes Screw and here's how we're getting on. 

Materials:
Wood
Wooden dowel (27mm Diameter)
1mm thick sheet steel
1.5m steel rod (19mm Outer Diameter)
1.25m plastic downpipe (65mm Inner Diameter)
Bracket for downpipe
Gearbox (We used one out of a cheap drill)
24V Motor

Tools:
Wood lathe
Metal lathe
Band saw
Sander
Plasma cutter
Bench Grinder
Angle Grinder
Handheld drill
Gas Welder
CNC Machine
Bench Vice
G-Clamps
Solidworks software
 

Before we could begin to manufacture the screw, we had to do calculations for the optimum angle of the screw to sit at and for the pitch of the blades.  For this we referred to 'The Turn of the Screw: Optimal Design of an Archimedes Screw' by Chris Rorres.
Our results gave us 40 blades with a pitch of 30mm and we kept our angle flexible at this stage.

http://www.cs.drexel.edu/~crorres/screw/screw.pdf

In order to make the screw more efficient, we wanted all of the blades to be of equal size and pitch.  To make the blades of equal size we simply cut a template out of wood by turning a cylinder of wood on a wood lathe, slicing it into discs then drilling through the centre.

For the pitch we created a template using a CNC machine.  We know that a lot of people don't have access to a CNC machine so the discs can also just be bent into shape during the welding process using a hammer.

Cutting
Each disk is cut out of a sheet of steel using a plasma cutter.  The template [made in Step 3] is clamped to the sheet of steel then the plasma cutter is run round the template cutting the disk.

Grinding
The edges of the discs are then ground to the right size using a bench grinder [our discs were 64mm diameter].  This also removes the burr left by the plasma cutter.  The surface of the discs are also ground down using an angle grinder to remove any burr from the surface and improve the aesthetic.

Bending
Each disc is then clamped in the CNC template [made in Step 3] in the bench vice to bend it to the correct pitch. 

We then welded each disc to the central pole using a gas welder, we found from making prototypes that the gas welder gave a much cleaner finish than arc welding.  This created the blades of the screw. 

The welding caused the central pole to warp slightly so we had to manually bend the pole back so that it was as straight as possible. This took some time!  We made supports for the screw to sit in so that it was held at one level and span it using a drill to see where the warping was, we then had to bend it by hand.

Also at this stage when we tried fitting the screw in the downpipe it was too tight a fit so we had to grind down some of the blades using an angle grinder.  To find out which blades needed grinding we forced the screw in the pipe and ran it off the drill, this made the screw cut the inside of the pipe and when we pulled the screw out, the blades that had lots of plastic on them, and therefore had been cutting into the pipe most, were the ones that needed ground down.

Traditionally an Archimedes Screw would rotate inside the pipe while the pipe stayed stationary.  However we decided, after a lot of testing, with our pump it would be more efficient if the screw was sealed inside the pipe and the pipe rotated along with it so really now it is a variation on the coil pump.

The screw fits tightly into the pipe anyway but in order to make it sealed we first had to run silicon sealant [any bathroom sealant should work, we used multipurpose silicone sealant from the pound shop] round the central pole so that there were no gaps between it and the blades.  We then put a generous line of sealant round the outsides of the blades and then forced the screw into the pipe and left it to dry over the weekend [we recommend leaving it for at least 24 hours].  This sealed the screw to the pipe and this way hopefully no water will run back down the pipe like it would in a traditional Archimedes Screw.

This step is really ongoing throughout the whole build.  We found it was essential to keep testing our screw to see if it would draw up water and early on we had some pretty impressive results.  Above is a video of our first prototype working whilst being ran off the hand drill.

The frame is a very simple 'L' shaped frame made from wood.  We wanted to make it so that we could adjust the angle that the pipe sits at whilst its in the frame.  To do this we made a simple square frame that would slot into notches on the sides of the tank we'll be using.  At the front of the frame are 3 notches on either side, a dowel with a bracket attached sits in a set of the notches and this can be moved, each set of notches giving a different angle. 
Two 'legs' come down from the back end of the square frame and this had two sets of hooks on it in which another dowel with another bracket sits in and this can alter the angle further.  The screw is held in place on the frame by a simple collar that sits over the bottom with a plate on the end that the end of the screw rests against and there is also another collar that sits before the first bracket and stops the pipe from slipping down any further holding it in place.

At the end of this project we have a competition to see which team can move the water most efficiently and in order to regulate the pumps all the teams were given the same motor [24V].  Now this motor produces 0.1Nm of torque and the drill we were using to test our screw produces 40Nm of torque so we faced quite a challenge to get our motor to produce enough torque to turn the screw.

We decided the best way to do it was to buy a cheap drill and salvage the gearbox from it.  We then attached it to the motor and we estimated this increased the torque to about 20Nm.  However once we'd done this we realised that by increasing the torque so much, we'd decreased the speed by too much and the screw was now going too slowly to pull up any water so we had to introduce another gear box, this time gearing it back up slightly.

We also greased up the gearbox from the drill to make it run more smoothly.

We had to find a way of holding the motor safely above the screw and it also had to be held in such a way that it couldn't rotate.  To house the motor we used the casing from the cheap drill as it was already the correct size so fitted perfectly over the motor and chuck.  We cut the drill handle off and sanded down the rough edges to get the right shape.

We decided, as we are Team Shark, as a bit of fun, to make the motor housing look like a shark so we first spray painted the casing white and then shaded a blue grey colour over the top.  We added an MDF fin and voila - shark.  As I said, this is just a bit of fun and it passed the time while we were waiting to get things done in the workshop.

The motor is stopped from rotating as it is screwed to the back plate of the holder and the holder will be securely attached to the frame. 

When we did the test runs of the screw, we found that when the water was exiting the pipe a lot of it was running back down the outside of the pipe and because it was still spinning when it exited it tended to splash everywhere.  So we decided to create a funnel to channel the water into the exit bucket.

Firstly we vaccuum formed over a hemisphere to create a bowl type shape.  Then we drilled and filed out a hole that fit tightly over the end of the pipe.  Finally we heated up the plastic again using a heat gun and bent it into the shape of a spout.

Sadly we couldnt get the final screw to run off of our motor but it would work fine with a more powerful motor or off of a drill like the one we've used for testing throughout the whole project.  Heres the final screw working off of the drill