fe1rx adds full aero

[lowside67]

Originally Posted by fe1rx

Getting data into AiM dash loggers is really easy. The logger provides the 5V source, and takes care of calibration, including non-linear calibration (as might be needed for a rotary position sensor). Unfortunately the Solo DL loggers don't accept additional inputs, unless they can be located on the CAN bus, and accessed by a custom protocol. I have had no luck with getting details of the BMW PT6 protocol from AiM, despite some prodding, so I suspect the details are considered to be proprietary. I think that means my next logger will not be a Solo II DL, but one of their dash loggers.

With regard to rotary sensors, AiM pointed me at the 2004-06 Cadillac Escalade ride height sensor. Various other GM models with continuously variable real time magnetorheological damping appear to use essentially the same sensor with differing brackets by application and widely varying pricing. The cheapest variant I found (Dorman branded) was only about $32 USD each at Rock Auto, making it very hard to beat, price-wise.

The mating connector (Standard Motor Products S1663) is also available at Rock Auto.

Motec repackages and uses a GM sensor also. They indicate that it is "contactless", i.e. a Hall effect sensor mimicking a potentiometer. Likely this is better with respect to signal-to-noise ratio than a true potentiometer.

https://www.motec.com.au/ac-sn-position/sn-position-rp/

I am also investigating the Mercedes linear pots. I suspect they will be a bit pricey though. Also, a lift gate normally works in a 0.1 Hz environment, whereas the GM sensor is intended for the suspension environment so it may be a better choice.

Depending on the number of channels needed, the AIM MXM is a fully featured dash logger with integrated GPS, 2 CAN networks, etc for $1100usd - good value when you look at the price for the next alternative up (MXL2). I have one and it works well, though I am considering upgrading for more screen real estate. While it can only handle 4 channels natively, I believe a hub can be added to increase this capacity.

The Mercedes R class rear tail gate linear sensor referenced in the Lotus thread reports that it can generate 1000Hz (!) results and he includes a trace from his datalogger, I don't think that'll be an issue.

With respect to your point on the connectors, those might be the matching connector for the sensor, but you'll need connectors to interface with the AIM harness unless you plan to cut them off. Newark sells the Binder 719 connector which is used on the dash loggers for like $10/each and then you have a full plug and play option without hacking up your AIM harness which retails for $350+. Just food for thought.

-Mark

[Nugget]

Originally Posted by fe1rx

]

I am also investigating the Mercedes linear pots. I suspect they will be a bit pricey though. Also, a lift gate normally works in a 0.1 Hz environment, whereas the GM sensor is intended for the suspension environment so it may be a better choice.

Further on in that thread he says he paid $20 for them and got them logging at 50hz after calibration.
I have no personal experience with them, just throwing out a potential option.

[rac]

The sampling rate is dictated by the logger, the sensors themselves provide analogue signals. From memory ~200Hz is the type of resolution that would be ideally targeted for track dampener / suspension analyses but 50Hz would still be valuable. Regarding noise I guess you just have to suck it and see, if you start with high sampling rates can always smooth and filter the raw data anyway, assuming you have the logging memory capacity.

[rac]

Originally Posted by Nugget

Further on in that thread he says he paid $20 for them and got them logging at 50hz after calibration.
I have no personal experience with them, just throwing out a potential option.

Great find. The sensors sold for this purpose are a huge rip off considering what they are.

[fe1rx]

2020 revealed issues with excess heat. Addressing those will require both heat hardening of some of the susceptible components, and reducing temperatures in critical areas.

Addressing hardening first, both the brake mid flex hoses covers and cover on the LH parking brake cable suffered from heat damage.



A quick test with a thermocouple and a thermostatically controlled heat gun showed that both the brake hose outer cover and the end boots how signs of degradation at 150°C. Techna-Quip will supply brake hoses with a fiberglass heat shield on request. This shield also has a metalized reflective outer layer to reflect radiant heat. These look promising. To address the parking brake cable I installed a fire-sleeve cover in the affected area. While this lacks a reflective layer, it provides substantial insulation.



My earlier attempt to increase the stiffness of my exhaust hangers to reduce the sway space it occupies was not successful. The urethane hangers were stiffer (at room temperature) but failed by melting.



The OE rubber mounts seem to survive the heat environment fine, so I modified a set by filling the inner cavity space with RTV high temperature silicone. Regular silicone can only be applied in ¼” thick layers as it cures in response to air and humidity. Beyond ¼” and the material will never cure down deep, so multiple layers and multiple days were required to fully fill the cavity with silicone. A 2-component activator-curing silicone would have been a nicer solution, allowing me to fill the cavity in 1 shot, but I didn’t have any of the stuff.





A quick test of the stiffness of a new OE hanger and a modified hanger to be approximate 4 times stiffer. I did this by hanging 5-lb and 25-lb weights vertically and measuring the pin-to-pin distance with a caliper. The added stiffness makes them a bit harder to install, but I am very happy with how firmly the exhaust is now restrained.



Temperature reduction will be the subject of another post.

[fe1rx]

In order to better manage transmission, exhaust and differential cooling air flows, I have added a baffle and some louvers. The exhaust baffle now separates the exhaust cooling air flow more completely from the air passing over the transmission. Here is the forward bay before adding the baffle, looking to the right:



and looking left:



Here is the view looking right, with the baffle installed:



and looking left:



The orange silicone strip is a close fit with the forward panel, and it prevents metal-on-metal contact when the forward panel flexes.

I was reluctant to install hot air outlets (i.e. louvers) in the undertray last year, because dumping low kinetic energy hot air under the car will reduce its effectiveness. At this point, I think that the undertray is effective enough that it can afford the loss, and experience has shown that I need to add some cooling. I still want to have the transmission/differential cooling air flow the full length of the undertray before exiting at the rear wheel wells and above the diffuser, but I decided to pull some of the very hot exhaust cooling air out with a series of 8 louvers that follow the path of the midpipe.

I was unable to find a commercially accessible louver that I liked, so I designed a die to press out what I had in mind. The elongation required to form the louvers is too much for hardened T6 grades of aluminum I have used in the forward and mid panels, so I made the louvers as separate removable items. As they are susceptible to damage, making them removable also makes that damage easy to fix by replacement.

The louvers are made from 0.063” annealed aluminum sheet. The blank is CNC drilled and slotted, with the finish shape engraved to make trimming easy. The blank below is an early trial with some cracks visible. Cracking was later eliminated by refining the die and the technique.



Once the blanks were trimmed and deburred, 10-32 Pem nuts were installed in each fastener hole to make the installation one-handed.



The associated cutout was made easy by creating a router template.





Installation is from the inside, as this eliminates an external lap joint to spoil the air flow.





Hot air spilling from the louvers will result in warmer air entering the differential NACA inlet than did prior to the louver installation. That isn’t really ideal, and if I have cause to rebuild my diffuser, I will consider moving that NACA inlet laterally to put it in cooler air. Before I worry about that though I will take some temperature measurements and do some flow visualization to see if this is actually a problem. The volume of air flowing under the car is so much greater than that flowing out of the louvers that dilution of the hot flow may make the issue insignificant.

Time for some testing.

[AndyW]

fe1rx,

Everything you make is a work of engineering art. If someone hasn't called "dibs", if you ever get out of the 1er game, I will drive to Canada and buy your car and any additional random parts you have...

[lowside67]

That's not a 1er - that looks more like an Airstream

-Mark

[fe1rx]

If we want to be able to characterize the performance of our aero mods, we need be able to measure two things:

- the downforce acting at each axle
- the dynamic pressure acting on the vehicle

We don’t have a wind tunnel, so our testing will be done on “the road”. With appropriate calibration, we can measure the weight on each axle, and by comparing it to the static weight on each axle, we can calculate the aerodynamic contribution.

To measure the dynamic pressure, we need a pitot-static tube. A pitot-static tube detects the total pressure (aka stagnation pressure) and the static pressure. The dynamic pressure (Q) is the difference between the two. Hence with a pitot-static tube and a differential pressure transducer, we can log dynamic pressure in real time.

Anecdotal discussions of downforce generally speak in terms force (lbs) and vehicle speed. I offer up “Porsche says the car (GT3RS) generates over 900 pounds of downforce at 124 mph and an astonishing 1895 pounds at 177 mph” as an example. These statements have the benefit of being in familiar units, but are fairly useless for analytical purposes. There is an implication that the numbers apply at sea level standard day atmospheric conditions and without the benefit of a headwind, but what about other conditions? And how exactly can we compare these values to other claimed values for other vehicles and/or at other operating conditions?

A more general view requires examining the lift equation, which is actually pretty simple:

Lift = Q x A x CL

Where Q is the dynamic pressure (which we can measure with a pitot-static tube), A is the frontal area of the vehicle, and CL is the lift coefficient. For a race car the lift force acts downward making CL negative. Strictly -CL is the downforce coefficient, but often the – sign is dropped and I will drop the – sign going forward.

The drag equation is closely related:

Drag = Q x A x CD

Where CD is the drag coefficient.

For that Porsche GT3RS I have found the following published figures:

A = 2.14 sq m
CD = 0.36

We can learn a lot from the claimed “1895 pounds at 177 mph” data point:

177 mph = 285 km/h = 79 m/s



QA represents the force we have to work with, and CL represents how effectively we make use of that force to create downforce. CD represents how much of that force appears as drag. For a car without active aerodynamic devices, both CL and CD are essentially fixed values for all speeds.

CL = Downforce / QA = 1895 / 1848 = 1.03

As we now know the values of both CL and CD we can calculate the resulting drag at any value of Q and its associated sea-level standard day speed.



CL, CD and A provide a useful representation of the aerodynamic performance of a vehicle. If we want to compare two different aerodynamic configurations it is much easier to have that discussion in terms of those three parameters than in terms of speeds and forces.

Anecdotal discussions about race car pitot-static tubes talk about them measuring airspeed. It is true that a pitot-static tube is used to measure airspeed in an aircraft, but the details are fairly convoluted. They measure indicated airspeed, to which a position error correction must be applied to get calibrated airspeed, to which a density ratio correction must be applied to get true airspeed. While a race car pitot tube will need a position error correction to be applied to get an accurate reading of Q, none of the airspeeds mentioned above have much relevance to a race car. As has been demonstrated in the above calculations, it is the dynamic pressure multiplied by the frontal area that is the really useful thing that a pitot-static tube can provide.