Velomobile body construction progress

First position of the diagonals and the wheel well jig

First position of the diagonals and the wheel well jig

The forward half leaving the floor and the tail. Shut faces are PVC foam.

The forward half leaving the floor and the tail. Shut faces are PVC foam.

The diagonals interfere with the drivers knees on the first trial fit.

The diagonals interfere with the drivers knees on the first trial fit.

Checking for sight-lines

Checking for sight-lines

The floor is supporting the weight of the body but there are many clearance problems

The floor is supporting the weight of the body but there are many clearance problems. The diagonals have been moved. The hinges are tried in place.

The shut faces have been replaced by plywood for durability. PVC Foam is stiff but a little soft

The shut faces have been replaced by plywood for durability. PVC Foam is stiff but a little soft

One of the front body hinges attached to a 3D printed attachment point

One of the front body hinges attached to a 3D printed attachment point

Frontal aspect

Frontal aspect

Velomobile: Zip-Tie Connections

test pieces velomobile construction

test pieces velomobile construction

The great advantage to using zip-ties for joining is they can be adjusted as you build. They are light and incredibly strong. I used them with PVC conduit tube and ply bulkheads to build the “”Blimp”. This is where I started again, this time with PVC tubing and PVC foam bulkheads.

IMG_2838

I knocked up a test to see if the foam bulkhead could take the strain of the nylon zip-tie which can be sharp. The tube took up a tilted position. I redid the test three times using different wraps until I got a self locking result with final straight pull through.

IMG_2841

IMG_2843

Using tubular stations brought another development line. Each solution also showed a weakness, and by trying to solve it, brought another knot. This is called ‘praxis’, learning by doing.

IMG_2813

The final process involves temporarily holding the frame together to get the spacing correct. A 5mm hole is drilled through and through the longeron into the frame. All frames and bulkheads are done at the same time. The longeron is then turned 90 degrees. Two zip-ties are used, starting on opposite sides, The 5mm hole in the frame is enough to locate the longeron. Using a single 200mm tie is possible, but it can be clumsy in confined spaces and time consuming. The extra block weighs 0.165gm. which is reasonable for all the avoided frustration.

Velomobile: Pixels to plastic

900MM STATION

900MM STATION FRAME

Once we had agreed the drawing Layout One.1 was a good starting point for a build, I started the process of converting the outline drawings to working ones. This means placing the 28 lengthways 15 mm pipe (longerons) inside the outer skin. I started with the 900 mm Station and made it the Master. This has a registration box that is used in each seperate Station, and remains identical throughout. My initial idea was to print the Station on to paper and use that as a template to cut flat boards. The registration box stops the programme deciding what size each one is best for you.

IMG_2570

Two things happened about the same time, I bought a laptop projector and Davy Jamieson introduced me to underfloor heating pipe. At 16 mm it has 2 layers like the PVC plumbing pipe, but it has an added layer of aluminium. When you bend it it retains the shape. I redid the drawings with 15 mm and 16 mm pipe and projected the drawing directly on to a board, taped in the registration box on the board and the feet of the Workmate and tripod on the floor

.IMG_2574

IMG_2576

IMG_2577

IMG_2582

I built a frame called a Strong Back to support and keep all the Stations properly spaced and in line. I should have stayed with my first plan of using aluminium extrusions. It is more expensive but it would not have twisted and warped like the timber in the centrally heated atmos. However this failure has lead to a solution which could give great benefits to structural frame strength and building simplicity, more later.

IMG_2598_2

IMG_2600_2

Charles and I then lashed the structure together loosely with zip-ties. At this stage nothing is really finalised, the drawing is one thing, what we have here is another and it can be useful go with what evolves rather than rigidly pursue an ideal.

Velomobile design: Iterations

LAYOUT 1.3.2RED OH

I made a start on a Sketchup 3D drawing to have something to point sticks and throw rocks at. The first task is to cover the rider and front wheels. I wanted to keep a simple shape, and deal seperately with the head above the shoulders to give minimum frontal area. The body frontal area is determined by the height and width of the shoulders, and the height of the knees as they cycle. The wheels cause a lot of turbulence. If you enclose them, then the body has to be wide enough for steering clearance at full lock. If you try to pursue one aspect of a design for purity of function it has a terrible habit of showing up flaws in another area. I don’t like compromise, but trying to find the best balance of the least offensive solutions is often the most you can hope for. I had already been through this with the “Blimp”, so that seemed a good starting point for development.

LAYOUT THREE COMBINE 1

Building the drawing in two parts and combining them was too much lke hard work and showed up how difficult the real construction could be. So I combined the head faring into the main body, at the cost of increased frontal area.

LAYOUT THREE COMBINE 1.5 nose

This looked OK until I put in the clearnce for the feet. This gave two ‘nostrils’ which appeared to be scooping air into the body.

LAYOUT THREE COMBINE 1.5

I added an air-dam but this unbalanced the look badly, I tried an elliptical dam guiding the air around the wheels and away again, but again it looked very difficult to build and keep light. The dam would have be able to to rise over Speed Bumps, adding complication, weight and jamming.

LAYOUT ONE.1

In order to divert the air around the feet I lowered nose, and the tail, to smooth the air flow and balance the shape.

LAYOUT ONE.1.1

LAYOUT ONE.1.3

Charles Barnard’s long distance cycling history

I wrote up the section on Charles’ cycling travels from memory and like a good journalist I got it wrong, here is the real story;-

Charles Cycling History

I retired in 2000 and began cycling as a hobby almost immediately, having done no cycling since the early 70’s.

In July 2001 I took delivery of a touring bike that had been designed to cope with my 6’4″ size and sitting upright riding stance.
I made several trips in and around Glasgow, and became addicted to cycling such that my distances rose from 5miles to 15, 35
up to 50miles over about 6 months. More miles seemed better and I soon went achieved 100miles, albeit at the pedestrian
average speed of 10mph.

My first long trip on the new bike was later in 2001 from Pollokshields through the city and up the Forth Clyde canal cycle path to
Bonnybridge. I took the road up to Stirling on to Callander then Crianlarich, where I turned to head back to home.
The trip was about 150 miles and took 13 hours to complete but proved to my satisfaction that the bike was ideally suited to
long distance touring. As usual I travelled at a steady 10mph.
I continued to cycle up to 100miles a day (3 days per week) for the new few years.

In Jun 2002 I loaded up the bike with panniers full of luggage and took the train down to Plymouth, where I took the ferry
across to Roscoff in Britanny. On this trip I travelled up to Rouen across country to Rheims, down to Troyes, Dijon and finally
arriving in Lyons having completed 1000miles.

In 2005 I undertook a charity event to cycle around the edge of France from Calais to Brest then Nantes to Bordeaux.
I continued down to the Pyrenees and then went from Biarritz to Perpignan in the foothills. I followed my Route up to Montpelier
on to Marseille then to N
ice. From here I followed the road to Grasse then Avignon and up the Rhone to Lyon.
The journey continued to Besancon Mulhouse Strasbourg up to the eastern top most border with Germany.
Here I turned west and followed roads via Metz Sedan Lille and back to Calais.
Total 4800km (3000miles) took 64 days in the saddle with 30 rest days. On average I covered 50 miles per cycling day.
Helen my wife drove our motorhome so that I always had a comfortable bed to recover from my ride.

I did encounter several saddle sores on route which made me think seriously about getting a recumbent trike,
as it doesn’t have a saddle but a proper seat. Another factor was battling the elements in an upright riding stance is very tiring.
The recumbent position doesn’t suffer from this problem as you ride in a laid back position with legs out in front.

In 2008 I cycled from Saint Jean Pied de Port over to Pamplona the across northern Spain via Burgos, Leon and Astorga to complete
the Pilgrimage to Santiago de Compostela ( St James Field of Stars) a distance of 500 miles. It was a very enjoyable journey.
In an average year 100,000 people walk the pilgrimage and about 30,000 cycle it.
Whether you are religious or not everyone is friendly and the welcoming. You carry a passport with you and get it signed along the way.
When you reach Santiago you exchange the passport for a certificate of absolution, or achievement depending on your beliefs.

I enjoyed the experience so much a repeated it 2010, and my trike broke about 100 miles from the end. I completed the trip in
Feb 2012 on my touring bike.

One feature of cycling that I found very dispiriting is adverse weather. Battling sting winds on a bike can result in almost no
progress with your journey. This battle is lessened by using a recumbent, but the cold and wet are bigger problems than conventional cycling.
An ideal solution for me would be an enclosed trike or velomobile.

This is what Kenny is busy designing and building.

Regards Charles

Velomobile start of a new project

Charles wants a Velomobile, a small human powered car, to complete a challenge of cycling round Britain. Charles Barnard is a long distance cyclist, he warmed up by doing Lands End to John O’Groats followed by 4,800 kms round the perimeter of France, then capped that by cycling  864 kms along the Pilgrim’s Way, up and over the Pyranees, to Santiago de Compostela, in Galicia, N.W.Spain. All of these expeditions have three things in common, weather, wind and sun. Singly and in combination these elements can be very wearying, to the point of psychological collapse. The answer, Charles decided, is a Velomobile. The shelter will keep you dry and the aerodynamics will reduce the effort required. If the reduction is only 10%,  it is cumulative, day to day, and then over 100 days cycling results in a saving of 10 days in simple terms. That is a lot.

 

CHARLES TRIKE LAYOUT 1
CHARLES TRIKE LAYOUT 2
CHARLES TRIKE LAYOUT 3
I was introduced to Charles through Walter Galbraith and the Outdoor Group, a walking and cycling club. I told him about my experience building recumbent bikes and aerodynamic Super Mileage cars. and I offered to design and build a body for his recumbent trike. I took initial measurements of his trike for clearance dimensions. The next day I sent Charles a couple of photos of my cars and also one of the construction details.of the Blimp. Nothing really happened for about a couple of months as Charles was, and is, tied up in a property refurbishment in the South. I prepared five 2D sketches laying out options and body plans, open and closed. One of the main topics of discussion had been that I thought Carbon Reinforced Plastic was difficult and messy to prepare, expensive, and worst of all, heavy. I’ll post on this later.
CHARLES TRIKE LAYOUT 4
CHARLES TRIKE LAYOUT 5
All this discussion sbout Velomobiles, got me thinking about building my own for some long distance touring. I resurected a drive system I designed many years ago but never got round to building. It is a Linear Drive which reduces the frontal area of the feet when pedaling. The design for this trike and Charles’ trike body are all based on the principles of using ellipses to generate volumes from one calculted profile to another. I worked up an illustration version to show the difference between that and a plain profile plot. The upper and lower body profiles are NACA 0008. The equator is NACA 16-021. The head faring is also NACA 16-021. All of the plotting co-ordinates are manipulated in an Excel spreadsheet.
CYCLECAR TRIKE LINEAR DRIVE 16021

CYCLECAR TRIKE LINEAR DRIVE 16021 ILLUSThe outcome of our next meeting was although Charles appreciated the 5 examples what he really wanted was my illustration version. I explained he had been seduced by a prettier picture and it is really unwise to set off on a trek round Britain in a untried trike. So I incorporated some of the shape ideas into one of the 5 and set to work building a 3D drawing in Sketchup, which I had only just started using.

 

RepRap 3D printing

Trying to describe what I wanted to do proved to be very difficult. I realised I needed a 3D drawing package, I asked a few people where to start to look. The range stretches from Free-ware to professional packages costing many thousands of pounds. The general consensus was to go for Blender, it is a package used for modeling 3D animated figures and rendering illustrations. The modeling principle is essentially creating a form in 3D space and then distorting it until you arrive at a result you can live with. I found it’s a very good programme, but not the way I wanted to work. I tried a couple of 3D CAD samples but they were massively competent but quite clumsy to handle, just too much.RepRap Prusa Mendel

At this time, a couple of years ago. I became aware of 3D printing. After a fair amount of surfing and reading I discovered http://reprap.org/wiki/RepRap A project started at Bath University under the leadership of Adrian Bowyer. It is Open Source and has had a huge influence in spreading the technology of 3D printing. RepRap stands for printers which can (self) REPlicate RAPidly. This is exactly what I was looking for. I could draw it in 3D, print it, develop the thinking and forms, redraw and reprint.

The process starts with a drawing made in Sketchup http://www.sketchup.com a free-ware programme which started life as an architectural CAD package and has developed into product design and sketching. The simple drawing is then transferred to Slic3r where the object is “sliced” into 0.25 mm layers, this data is then transferred to Pronterface where it is converted into G-code to instruct the machine on how to move and with the correct amount of plastic. Sketchup is a great programme because it appears simple, very straight forward and accurate. I regularly work in 3 decimal places of a millimeter, that is sufficient for what I want to do at the moment.

IMG_2608

I bought a Prusa Mendel machine kit and built the mechanical hardware. I needed help with the electrics and a friend Walter Galbraith, a lecturer in Electronic Engineering, came and sorted out the wiring. I don’t speak electricity at all. It started as a hobby, building and learning about this new machine and getting it running. I was working on another project and realised I needed a clamp to hold some tubing in place. I measured up the pipe, drew it up in Sketchup, put it through Slic3r, Proterface and started to print in 17 minutes. The end result is EXACTLY what I need, not close or just about, but exactly what is needed. At this point the printer changes from an interesting past-time to an essential tool.

IMG_2576_2

Developing Drawing Volumes

Images for this article can be found in First Prinipals Gallery

While I was delving into the surface of the gigantic field of aerodynamics I came across some figures relating to fish. Birds and humans do a very good job of propelling objects through air, but fish outstrip anything humans can do through water. Hydrodynamics is basically the same problem as aerodynamics, it is just the medium is so much denser and cannot, in any real sense, be compressed. Nuclear submarines with all their gigantic reserves of power can acheive +/-50 knots submerged, or that is what the owners admit to. Sharks have been measured at 55 knots and Blue Marlin have been measured at over 60 knots, admitedly over short distances. Both of these animals have been developing their forms over tens of millions of years. So what is it about their shape that allows them to do this? The first thing that struck me was most fish are not circular in cross section, but elliptical. Now this was not an easy fact to find as in all the web searching I did, I found masses of material, images, giving the distinguishing features of each species but they are all in profile, side-on. Nothing on their cross section. I knew fish are elliptical but I couldn’t prove it. How strange is that?

NACA 63-212 FIRST PRINCIPALS

So…..using my trusty 2D drawing package I started to try and generate a fish shape. I took a NACA 63 profile and at each station I drew an ellipse at 0.75, that is the radius is three quarters the radius of the circle at 90 degrees. (Stations are at right angles to length ways axis of a boat or a plane and become the cross section). I then added radii to mark out where the length-way pieces would go, they are know as longerons. In exactly the same way I did with the tool drawings I projected the intersections back to the vertical and horizontal.

 

I created an elliptical form in three ways;-
By drawing an ellipse using a constant ratio.
By manipulating the co-ordinates in Excel directly at 0.75 and plotting the result.
By applying the formula for an ellipse to the co-ordinate data
The drawing process was by far the simplest and quickest method and for this theoretical phase was accurate enough.
NACA 63-212 FIRST PRINCIPALS 1
NACA 63-212 FIRST PRINCIPALS PLAN
Why bother with all this, why not just draw a line? I had learnt the quickest way to develop something new, and understand it, is by small incremental steps altering only one aspect at a time. If you attempt two or more aspect changes at once it appears to be faster, until something goes wrong. The time required to unravel what has happened and rectify it is much, much longer than if you stick to the pedantic, plodding course. I also did not have anyone to ask, maybe they were there, but I could not describe what I was attempting to do, clearly enough, to ask the question.
So I now had a fish shape, complete with lines, where I could see the form if I married two drawings in my head. I realised that what I was seeing was a shift in shape. If you take a profile and rotate in a circular fashion around the central axis the shape remains constant, the shape at 90 degrees is the same as the shape at 0 degrees.  Now if the same shape is rotated around an ellipse the initial profile is the same as the circular path one (90 degrees) but as it moves along the ellipse it is part one and part the final shape at 0 degrees. At half way it is half of one profile and half of the profile at 0 degrees. As it continues around the complete ellipse it returns back to the initial profile.
The next step was to reverse the process and increase the ellipse to 1.25 of the original. This worked just the same, surprise, surprise. Add fins, tail. eyes, bingo, a symmetrical fish. Very few things in the natural world are symmetrical.
Now this is where it starts to get complex and it is why I was taking such baby steps and I was feeling my way along. I then took two different profiles that had co-ordinate stations which coincided and drew an ellipse between a profile at 0 degrees and another at 90 degrees.
The ratio of the ellipse changes for each station. Using this I connected three different profiles. The upper profile has it’s maximum depth near the back. The central one, the Equator, is more evenly loaded, and the lower profile is front loaded. Each of these profiles in their own right are calculated for minimum drag. The leap of faith I am making is that the shapes generated between these profiles are also minimum drag, although this is by no means a certainty. The resultant shape generated using this method looks remarkably like a Blue Whale. All of the co-ordinate points to make this form are numerical so the shape can be made using CAM (Computer Aided Manufacturing), this is very important.
MULTI PROFILE FISH 2 NPL9510
MULTI PROFILE FISH 2 NACA5H15
MULTI PROFILE FISH 2 NACA63012A
 MULTI PROFILE FISH 2 STATIONSMULTI PROFILE FISH
The final step in this series was to make a shape using four profiles connected by ellipses. If all the profiles are different then the resulting form is assymetrical and looks like your lower leg with the shin bone and three different outlines of the calf muscle. All of this work was done with an ancient version of a simple 2D drawing programme, no 3D work, no Computer Flow Diagnostics, just simple thought process and first principle thinking. I had come to the point where I could ‘see’ what I had generated but I needed to be able to show it to other people in 3D to make it clear.

 

Drawing volumes

2.0 Radial spacingShortly after receiving a drawing program, Paint Shop Pro, I had to come up with an  illustration of the concept for an oilwell tool. I had no one to ask, so I had to come up with my own solution, which entailed viewing the subject from above, in plan view. I then drew radii and where they cut the circle I projected them back to the ‘horizontal’. When the subject was rotated, 90 degrees, to a side elevation, the edge of the facet was then measured out. I had no idea if this was correct from a Technical Drawing or Draughting practice, but it worked well enough to give a 3D appearance to a 2D drawing. By adjusting the diameter and connecting facets I could create cones, recesses, chamfers, etc. By using layers I could alter the tone value of a flat colour or textures, using the guidelines of a template placed above it. This was long before 3D programmes do all this for you automatically.2.0 Taper Mandrel

Using this simple drawing procedure I started a series of sketches for ceramic pieces. They were based on the basic language of down-hole tools. The langage first comes from the need for strength, that there is enough material to stand up to the forces, stress relief to make sure it doesn’t snap and then ease of manufacture. There is another aspect which tells you on the outside how it is put together on the inside. They are purely functional pieces of equipment, they have no aesthetic qualities applied to them, they are not styled in the least. Consequently these tools have an appearance and form that is totally their own. I first had to build the machinery to make them, which I did eventually start, but circumstances overtook that project

6-1-4-conn-queen-body-structure

I did keep the idea of portraying an object by rotating from the profile to the cross section and back again. . The first project to fully use this was the Super Mileage car which was nicknamed “The Blimp”.  I drew up a NACA 63 profile and scaled it until the shoulders of the driver fitted inside and the rear wheels also. The chassis was divided into bulkheads to carry loads from the wheels and the drivers weight. The upper body was divided by into stations. This is how boats and planes have been designed for decades. The chassis is an extrusion of the NACA 63 in that it is straight sided and is the same cross section looking from above, it is pulled up. The upper body is the same profile but rotated 180 degrees, it is spun along it’s length.ScanBlimp Xsection

The head faring is NACA 63 but rotated and scaled to allow the driver to turn his head, and still keep the minimum 4:1 ratio, it is 4 time longer than it  wide. It is both spun and extruded. A lathe works by spinning material and applying cutters to achieve the profile. A mill cuts depths vertically  and horizontally and can achieve an ‘extruded’ result. This is how the 3D drawing programmes I have used work. Draw a profile, spin it and it looks like a lightbulb, or a wheel, or a bowl. Draw a circle and extrude it and it becomes a column, a rectangle and it becomes a house brick or a table top.Second attempt3

 

 

BOC-1(36)By using a calculated aero profile like the NACA 63 I was trying to give myself the best chance of a aerodynamically clean shape without having to pay for wind tunnel time, I did not have a Formula 1 budget. These curves have a tension that comes from guiding the path of the air as the object moves through it. If the curve is not right it stops guiding and starts disturbing the air until it can be as bad as if there was nothing there at all.

The design of the gH2Ost, the second Mileage car I built, was a natural development of the first. The sides were a profile, MH-27 rotated 180 degrees. The central profiles were scaled to give enough volume for the driver. The head faring to the the tail was a third profile. All three profiles are low drag, calculated curves. They were then connected simply, hopefully creating low drag shapes in between.
The development of this idea became a project on it’s own……..

The beginning of a new direction

 

 

Event guide

It’s strange how some things begin. Davy, my very good friend, told me about an event he had bumped into at the weekend. He had taken his boys to see the Shell Eco Marathon run at the Alford Motor Museum. Small racing cars drive round a track and the team using the least amount of fuel wins, easy. Or so it appeared until I put in a little research. It turned out this was the tip of an iceberg involving 128 competing teams from schools, universities and individuals spread right across Europe, Scandanavia, USA, Japan and Australia. Shell had started the challenge and had sponsored the event for more than 30 years. The main attraction for me was there are very few rules. The vehicle has to have more than 2 wheels, it has to have a human driver in control and it has to move under it’s own power. There are safety and construction rules to specify mirrors, seatbelts, rollbars, fire extinguishers that sort of thing. If you think you can win by using an engine from a double decker bus, try it, find out.

Team Fancy_Carol

I researched the most successful teams and the cars they had built. They were all very small and very light with small and light drivers. Aerodynamics seemed to play a major role which is surprising for vehicles which move so slowly. The engines were comparatively small, but there was very little information available. I have always enjoyed building light weight structures and using light weight materials. I knew I could not compete building an engine but I could by designing the chassis, body

Initial sketches

and geometry. Most of the cars were Tadpole tricycles with the 2 front wheels steering and the single rear driving. The driver lies prone with their head tilted just enough to see over their feet, this gives a very low frontal area. Mostly the front wheels were encased in faring front wings. Rightly or wrongly I figured this would lead to the air travelling into the tunnel between the farings would increase in pressure leading to turbulence and drag. I would attempt to divert the airflow as much as possible to avoid barriers.

I drew up a series of layouts gradually narrowing down compromises until I was happy with the overall scheme. After a short period I returned to the design and it struck me what do these lines, these curves, actually represent?, Initial drg1.pdfand can i do better. I started to delve into aerodynamics, especially concerning Human powered vehicles (HPV’s) and World Solar Challenge cars. Both have very limited power available and both have made incredible advances in performance.

At the heart of the aerodynamic work are databses containing NACA airfoil profiles. These profiles are calculated, not drawn and have been used in mostly aircraft applications for wings, struts, propellers but have also been used for designing boats, rudders and keels.

I downloaded a profile (NACA 0010) ready to copy the curve and away we go……. NACA 0010.dat

 

What I found was two lists of numbers, six digit decimals. I was not expecting this, I was expecting an image. The numbers are X and Y co-ordinates. This was the begining of a line of research and work that would last for years, until the present day.