Flow Simulated Tests

I have found a nifty free program that could do a flow simulation and show it graphically as a video clip in .avi format.  It is called Flow Illustrator.  It is not accurate to replace a full blown Computational Fluid Dynamics software but it is free.  You will need to do your own model.  This is simply done by using a side view of the model.  You can do this by using an existing picture with non essential details such as background and people deleted.  I use Microsoft paint to edit my picture.  Please note that it is advisable for you to use the CUT function using SELECTION.  Once the image you need is cut from the original, you crop the rest leaving with a totally white background.  You can upload the picture and set three (3) variables; Reynolds number, time between frames and video length in seconds.  I only change the setting for video length.  If you want to do comparison of flow between two designs, you need to have the picture in the exact dimensions.  The value of the Reynold’s number will be shown on the video.

You can see the comparison video on this link.

A screen capture of the comparison is shown below

SNR3 with Deflector

Simulated flow on Satria Neon with Deflector


Simulated Flow over Satria Neo with Standard Spoiler

Notice I have cut out the wheels to indicate flow of air under the floorpan.  Green coloured areas indicate high pressure/low velocity and Red indicates low pressure/high velocity.  The comparison helps to show that the original design with standard spoiler has a bigger wake left in the trail of the car.  The one with the deflector has a smaller wake as the height of the turbulence is reduced.

If you do another comparison with a sedan design, you would see that the wake is about the height of the boot.  The deflector now looks like what we intended it to be.  It is to reduce the wake by delaying air separation from the hatch.  This partly answer the previous question of why the Fuel Consumption curve is not logarithmic.  The deflector has therefore made the car more aerodynamic.  It makes it easier to cut through the air.  I would not be surprised if it has reduced the Cd value of the car.

Making the car aerodynamic has the added bonus of reducing drag and obtain faster acceleration and top speed.


Real World Test

It came as an afterthought.  It was never my direct intention to have done what I made.  I had initially made a Secondary Intake for reducing fuel consumption of this car.   See my previous article.  Then I wanted to make the spoiler adjustable so that it can actually be useful.  This was on my previous thinking that the spoiler functioned as perceived.  That it provided down force and that if I took it out, the handling would suffer.

This then changed somehow without any premeditated intentions.  What I have right now is a combination of efficiency solutions.  The secondary intake initially showed that I could save money in the movement of the throttle.  That was fine.  Then the aerodynamics were sorted.  Never did I thought the two combined mods made significant fuel saving changes.

I did a test run yesterday for 230km in total.  This was a return trip on the North South Highway.  It was to test the Deflector at speeds above 90km/h.  It performed admirably by registering an Indicated average Fuel Consumption (FC) of less than 5.0 l/100km. You can see the YouTube video here . Yes, this means I now have a YouTube account.  Forgive me I am rather slow at 46 years old.  I have only now found the use for it.

During the same journey I made similar tests for speeds of 100, 120 and inadvertently 130km/h.  All of these test results were on solid straight roads with the AC switched on.  I plotted the values which I had observed on a spreadsheet.  I had expected a exponential curve from 100 to 130km/h.  But I could not get it.  The only time I observed a high FC is when the car was being driven at 100km/h up a 20 degree incline.

capture_11092014_004322The table is now UPDATED on 9 November 2014.

The data is based on observed data from factory standard equipment.  They are not absolute and the value is AVERAGE Fuel Consumption.  The last data if drawn to scale should be closer but it should also be higher.

A few things floated in my mind.  As engineers we are told that aerodynamic load is very much affected by speed of the object or vehicle.  This aerodynamic load increases .  Fuel consumption will increase dramatically as speed increases.  Perhaps it is too early and too easy to say that somehow with the combination of both modification, we have come to a point where the world is no longer flat.  I might need to spend more day trips to Melaka to do more tests.  One might say that these values at these speeds are insignificant as the speeds are low.  But these speeds are what normal people are driving at.  This is the whole point of doing these modifications.  It is so normal people could save money driving normally.

Should anyone ask about performance degradation; I simply did not see any.  The maximum top speed is still attainable.

Deflector Part 2

Just a quick one as I do not want to give away too much.  Finally managed to do up the fibre glass work with Chooped Mat and Roving.  The Roving is better suited to hold the surface on the edges as it is stronger.


Producing the top surface Needed to brace this securely on this working design.  I would not need to get a really solid grip on the production model as it will totally be fibre glass.  As it is now, wood is used as the main structural material.  I would only need this working design to make some sets of mould.  The finished product would therefore be much lighter and look nice.

Deflector mounted on car

Deflector mounted on car

Had a spin around Putrajaya.  About 45km distance.  First round registered 4.3 l/100km while the second registered 4.5l/100km.  Will keep doing test and maybe try to add more surface area on the bottom.  I am very happy as it is, there was little effort to try and clean it up.  Not even fillers and it still meets the target.  Probably I might get better results with the extension.

Adjustable Spoilers Part 5

A day after posting Part 4.  I did a calculation using a spreadsheet.  I could either stick to my existing HITEC or get a slightly bigger one.  There’s a local Arduino reseller that has lots of TowerPro MG945 which has a stall torque rating of 12kg/cm at 6VDC.  I am definitely going to buy this.  Although the gears are of metal, due to the power that’s required, I prefer Nylon to take in the additional service demand.   I would also need to reduce the surface area by 30% while limiting the angle to not more than 30º.  As preparation to accept this servo I had to install a 3 wire cable taken from a power cord extension.  The length is about 1.3m which is sufficiently long to have the new controller next to the mid console close to the handbrake.  I have routed this through holes that were meant for the rear water jet for the wiper and through one for the spoiler mounting.  It’s a bit harder than just that, as I had to remove two panels and I had to make sure the cable was not interfering with the rear seat belt mechanism.

Cable Routing for Servo

While at the same time, I found the bracket for the reverse sensor a bit too big for my liking.  I therefore made holes to shave off 20g.  It was not planned.  Since it was hard to remove these panels, I would not be seeing it for many years to come.

Reverse Sensor BracketA few more days and I finished my mechanical build of the spoiler and had it fitted on the car yesterday 27 May 2014.  I had a spin in it around Putrajaya, in the most decent of 80km/h speed.  It did not fell off even though all of the wooden part was held in place by nothing more than Dunlop Contact Adhesive and two screws.  I had another spin today at much more livelier speed around the same track and it is still there on the roof.  I might add more screws to keep it all sane.  I used normal green coloured Garden Wire to secure the angle of attack.  I do not want the angle to be too big that it would add drag and load onto the new assembly.

Rev1 Front 3/4 view

New Wing

Rev1 Rear 3/4 view

I am happy to report that on the first test, I recorded an 18% fuel consumption reduction.  Initially the average at 80km/h is 5.3l/100km.  Now it is 4.3l/100km.  Tell me again how much that hybrid costs!  This is amazing.  Previously I could only make 4.0l/100km on a flat straight road.  But with this spoiler. I could drive it normally without having to use hyper-milling techniques to get close to that.  It goes to show that it does make sense to make the air separation as late as possible.  The angle of the spoiler closely matches the angle of the hatch.  I cannot be too sure if I had also reduced the wake.

With that I have completed the mechanical build and managed to clean the wet kitchen area of my house clear from tools and debris.  This gives much joy to my wife.  The next steps will be to finalize the electronics.


Adjustable Spoilers Part 4


I promise that this is going to be longer than a Star Wars trilogy.  Best be likened to a Tolkein.  Much so like Elder Scroll’s: Skyrim the PC Adventure game.  A continuous journey where you pick up little treasures along the way to and learn new skills to make you stronger, faster and befitting to yield that shiny Elven Armour newly crafted with ores you found in a deep mine.  If one is to find an ending to this story, then it will take time as I have realized a few months into this project.  Every inventor ends up making something.  They probably will turn it into a product for most people to use.  But the treasure for the inventor lies in gaining more knowledge and experience.


New tools

I think I have mentioned before that by education, I am a Mechanical Engineer.  However, this was in the 90’s.  There was huge development in Malaysia to get it gears running.  It was so huge that the country could not get enough talent to fill in the many spaces required.  I had secured a job months before getting mu Diploma. Malaysia’s  Development rapid growth went right up to end of 1997.  In that span of time, one gets to fill up those voids fairly easy once one has accumulated enough experience.  I went into Control Systems, behemoths like Distributed Control System (DCS) and Fire and Gas Detection Systems (GDS).  Right up to the end of my current employment, the process of learning never ends.  Such is the immense volume of knowledge to be mined from this world.

I have now taken up Electronics as a new field of self study.  Prototyping an Arduino micro controller was the door opener.  That is formal engineering.  But learning how to solder is enhancing skill.  I have now acquired basic skills to properly use a soldering iron.  I have now accepted that Tinning the tip of your solder and keeping it clean is very crucial to get the heat on the pad as soon as possible.  The internet helps.  Not just Youtube.com but also forums for hobbyists.  This just made me realized that the older technology is very much relevant in present day.  Data mining is the process.  Before you mine you would still need to do a geological survey of the lands.  That geological equipment is in your brain.  How does it determine?  Some say through experience.  Some would say faith.


It has been some time before I had any update on this blog.  Much of that time was spent on developing the prototype.  On a daily basis my head is torn between making JUST a prototype where tests can be done on a raw specimen, to actually developing that product into something much closer to production.  I had the original idea to move the spoiler directly from a small RC Servo.  I have grown to appreciate this invention.  It is small yet simple.  It needs just three wires to operate.  In that time to appreciate, I have also come to know of its inadequacies.  Size is important in this project because I need to squeeze it into size predetermined by somebody else who has his or her own design constraints. Size is proportionally related to how big a torque it can generate.  Shall I make it anew or continue on this road to hack an existing spoiler?  Another fork in the middle of the road.

The original idea of directly actuating the spoiler had to be abandoned.  This is due to the design of these servos.  Where the environment this unit has to operate, it would face maximum wind force when the car travels at its maximum speed of 210 km/h.  I have to move the servo away from being the only source of mechanical support to the spoiler.  It is now hinged by aluminium tubes.  Now we look at the force that the servo requires to generate.

The force on a wing is calculated as

F = 1/2 x ρ  × v ² × Cd × A


F = Wind Force in N

ρ   = Density of Air (typically 1.146 kg/m² at 35ºC)

v  = speed in m/s

Cd = Coefficient of drag

A = Surface area in m^2

One thing that we need to understand in this equation is that the Force required to overcome air resistance is heavily influenced by speed.  Now we have to look at the surface area as one that is directly being influenced by the moving mass of air.  Therefore, if the surface is at an angle to the air, one has to calculate this based on the degree.  Based on an freely available educational information, we know that the drag coefficient Cd;

Cd = 1.28 × sin (a)

Where a is the angle of attack or inclination of the surface.  Plus as in trigonometry, we know that the maximum surface area is one where the angle is at 45 degrees.  Any larger, then the Cd would need a different number than 1.28.  We also know that the surface area is a plank with the size of 715mm wide by 90mm long.  Having an area of 0.06435 m²

Now we have the calculated force

F = 1/2 x ρ  × v ² × Cd × A

= 1/2 (1.146) (200 x 10 / 36)² (1.28 sin (a) ) × A

= 2263.7 × sin (45) × A

= 1600.68 × 0.06435

= 103N

Or approximately 10.51kg

This shall be the design constraint for the servo motor.  We need one that can at the very least, generate this much amount of force to be able to move the spoiler.  The HITEC HS425BB has a stall torque of only 4.1kg/cm at the maximum power of 6V.  The servo arm that fits our requirement based on location of the servo and distance to the control arm is 30mm (3cm).  Therefore, the available force from this servo at that distance is 1.36kg or 13N.  If we choose to use the same area of spoiler the capability of the spoiler is limited by

a)  Speed i which it operates and

b) Angle of attack of the spoiler

We would need a servo that generates at least 31kg/cm or 431 oz in or multiple servos just to move against that amount of wind.  Therefore, we have come to another fork in the road.  Single or Multiple?  Looks like this is going to be costly.  This then brings me back to where I wanted to do this project.  A variable angle of attack spoiler meant for reducing operational cost.  It was for the purpose to make life a little bit easier for normal people.  We would not expect them to be running at 200km/h.

Assembly Unit

While we leave the issue of the servo, we look at the other design aspect.  A few days were spent on trying to shoe horn pieces of wood into the current design.  It is difficult but I have found new respect for using wood as building material.  This made me look back at history.  We had woodwork classes back then in the mid 80’s.  We had good teachers.  May they be blessed for passing on the knowledge that I am using now.  Its back to buying wood craft tools such as chisels and planers for me.  A worthwhile investment.  Every stroke of the plane will be followed by gentle fingers flowing down on the grain.  Feeling for inconsistencies.  You cannot escape from it.  Every push of a file, will be followed by a stroking hand.  This is why people were called craftsmen.  It wasn’t for the product they made.  It was for how they made them.


OLYMPUS DIGITAL CAMERAOne great advantage of using wood as a material is the ability to reduce the weight.  Prior to this, we measured the Fibre Glass material and it weighed in at approximately 1200g.  The wooden assembly itself registered just north of 900g.  It begs to question the need for composite material as a preferred building material.  In Malaysia the Forestry Department makes it a must for loggers to replant the trees.  It is a sustainable building regime only if everybody abides to it.  I would guess that the developed countries find the material takes longer processing time and have that added inconvenience of replanting trees being a requirement.



Further development goes on as with any challenges in life.


Adjustable Spoilers Part 3

I have purchased an Original Copy of an R3 rear spoiler for the Satria Neo to work on as a working development platform.  It is made out of Glass Fibre.  I normally would not do this if not for the fact that I was the told that the real deal would take some time to be ready.  I had gone to the Proton Parts Centre in Glenmarie, Shah Alam, Selangor to buy some parts for my other car.  I had asked for two sets but was told that it is not normally held in stock.  I was told that these parts do need to be back ordered.

Since what I am trying to do is make a full scale mock up and as a real working unit for further tests and development, I guess it would be ok.  It costs RM 140.00 which I purchased from an accessory shop called Brother’s from their main office in Kelana Jaya, Selangor.  There are actually two models of similar looking spoilers.  The difference is on the mounting holes.  I bought the proper one and took two more days before doing something about it.

I transferred dimensions from the spoiler to the wooden prototype wing and vice versa.  The wooden wing was wider than the spoiler’s trailing edge.  I chose to cut by 40mm offset from the shoulder of the trailing edge.  The hardest part was trying to cut it exactly straight as the spoiler is curved on all axises.  I used another straight wooden piece to make it straight.  I used a wooden saw, a dremel and a powered jigsaw.  And it came out shitty.  In retrospect, I should have just used the jigsaw.  I just had to be steady with it as an oscillating power tool is difficult to hold steady.

Severed Wing

Severed Wing


Once the trailing edge was cut, I tried placing the Standard Hitec Servo Motor around the cavity to see if there was room.  This has been the main question in my mind for weeks.  If the servo was too big, additional modifications are needed which would drag my schedule longer.

Servo Fits!

Placing the servo for fit testing

You can see from that picture that the Glass Fibre wing was made in several separate pieces.  Where they mate and glued to each other, there is some internal projections that hampered the fitment.

While finding placement for the servo it dawned on me that it would be easier to use the glass fibre cutoff as the moveable wing.  All it need would be to fit wooden pieces at either end.  These wooden pieces would then act as a swivel point on one end and a driven end at the other.  The driven end would only be a place for me to install a Servo Arm.  Then it occurred to me to measure the weights of the wooden wing and cutoff to compare.

Weight of Fibre Wing stands at 650g

Weight of Fibre Wing stands at 650g

Wooden Wing

Wooden Wing weight is at 442g

This appeared to me as simply amazing find.  While it is well understood that Glass Fibre is light and commonly used as lightening body shells, in this case; wood would be the ideal material.  Wood has very good compression properties.  In a spoiler, there would be some downforce on the top surface and wood would be excellent in this aspect of design.  It is strong while remaining light.  The negative side of it is its long term condition when exposed to wet, hot and humid weather of the tropics.  A nice light coat of shellac or even Plastidip could help reduce that.

While I had the wife’s baking scale out, I weighed the remaining part of the glass fibre wing. It weighed in 1073g.  In total the whole glass fibre wing would then be in excess of 1.7kg.  If we replaced that with wood, we could see a weight savings of more than 20%.



This could lead to replacing most of the structure into wood.  Saves fuel consumption by making a cleaner aerodynamic profile while saving weight!  Looks like I am onto something good here.

Adjustable Spoilers Part 2

It has been a few weeks of browsing through my den of broken things to put together a test rig.  The reason being to please my self doubt about what I am trying to do.  A lot of times I have questions floating in my head if I could actually pull this off.  I have been messing around on the internet and trying to find something that could show that my half baked ideas is actually worth the time.  Time is on my side.  I have stopped working for the moment.  I ended my previous job as a Manager for Fire and Gas Detection system with Draeger Safety with a small compensation package.  That bit of kit would probably last me a few months before I actually need to find a job.  At 45 years of age, that is not going to be easy.

I have taken a piece of wood from my hoardings and bought a cheap set of planer.  The last time I used this was in Secondary School in the early 80’s where we learned woodworking.  I therefore whittled down this piece of wood for two days to come as close as possible to the profile that I needed.  I also used my dremel and coarse grit sandpaper to give a good smooth finish.  Woodworking is very addictive I must say.  I find myself constantly reshaping the profile untul it looked satisfying to my eyes.  I was not looking for an effective aerodynamic profile.  I wanted it to look good.  I had to force myself to stop because I needed to do this rig and reassure myself.

I took the finished piece of wood and proceeded to find the center of gravity.  This is crucial as this would help to reduce the torque demand on the servo motor.  Placing the servo in line with the axis of rotation means reducing the weight that had to be moved around the axis.  I simply held the spoiler by the side of both hands.  I slide both hands to the center of the piece.  Where the tilted, the weight has gone the other way.  I slid the hand which had no resistance because of this shift.  I put the wood on a pen and tried to balance it as much as possible.  This way, I found the center of gravity and marked with a pen.  I then used the grain of the wood to show me the pivot point at each end of the piece.

Then I looked inside the cabin again to find unused TV stands meant for my 42 in plasma TV.  I drilled both at an approximate height.  The undriven pivot side was held by a simple screw.  The driven side had a servo hub which was nailed.  The hub has a 24T spline which would normally be attached to the servo shaft and screwed in place.  This is another issue that I would need to address.  I would need to cut an approximate width of 20mm to enable the screw to tighten the sawn wooden piece against the servo.  Once that servo is tightened, it would then be screwed to the main piece.


Finished Test Rig

I have been trawling the pages of Servo City for a few weeks now and have ordered some bits and pieces.  I will tell why I need those a few paragraphs later.  I have made a test rig to show myself that the small servo could turn the wooden spoiler.  A video of which can be found on my facebook link here


Close up of Hitec HS-425BB RC Servo

What I have realized after playing with the test rig is that the weak link lies in the connection to the HITEC Servo.  This HITEC HS 425BB was bought from a local Arduino authorised agent at MyDuino.com.  It was the biggest RC Servo that they had.  I could see that at a speed of 200km/h and with a surface area of 7,500mm2, the wind load on the spoiler would only be contained by the servo hub and the swivel joint.  Even though this HS 425BB is a model with bearings, I would think that it would not be sufficient to handle the load. Therefore I decided a redesign is in order.  Part of the items I hard ordered online was an aluminium arm to connect to the spoiler.  But because the spoiler would have a different pivot point and than the servo, the hole on the arm must be modified to a slot.

Having two support on the driven side would increase load capability.  This also means is that the torque demand on the servo would be increased.  I am prepared to make the spoiler lighter by hollowing; making more holes where necessary.