What's new

excellent website for data on any car

freddofrog

won 16.4k on Euro lottery :)
Messages
7,326
Reaction score
847
Location
Off abroad
Car
TBD
I found a book in the house of my (recently deceased) father (a Chartered Mech Eng).

It had a lot of info on the "coastdown" method for finding the coefficient of drag (Cd) and the coefficient of rolling-resistance (Ad + Bd)

I had an idea that maybe we could all try some tests on each of our Accords, and determine if there was a relationship between the coefficient of rolling-resistance and the mpg that we get.

To begin with, I collated the 7th gen data from the 7th gen ESM
DIMENSIONS CL7/CL9/CN1
Overall length 4665 mm
Overall width 1760 mm
Overall height 1445 mm
Wheel base 2680 mm
Track - Front 1515 mm
Track - Rear 1525 mm

DIMENSIONS CM1/CM2/CN2
Overall length 4750 mm
Overall width 1760 mm
Overall height 1470 mm
Wheelbase 2720 mm
Track - Front 1515 mm
Track - Rear 1530 mm

WEIGHT CL7/CL9
Curb weight - 5MT 1301−1418 kg
Curb weight - 6MT 1386−1450 kg
Curb weight - 5AT 1357−1483 kg

WEIGHT CN1
Curb weight 1463−1533 kg

WEIGHT CM1/CM2
Curb weight - 5MT 1462−1545 kg
Curb weight - 6MT 1508−1582 kg
Curb weight - 5AT 1494−1610 kg

WEIGHT CN2
Curb weight 1575−1647 kg


Using the above data for my car, I did some coastdown tests to get the Cd and (Ad + Bd) but I found it very difficult to get a consistent Cd .....which should not change at all, and since I didn't know what it should be, I decided to search online for the Cd values and I found some values on some sites.

for 7th gen saloon (e.g. CL9), the "claimed" Cd = 0.26 ....but wikipedia has an excellent list with a more believable 0.30 --> https://en.wikipedia.org/wiki/Automobile_drag_coefficient#Typical_drag_coefficients

for 7th gen Tourer (e.g. CM2), there was an "estimated" Cd = 0.34 ....I can get close to this in the tests using my car

By retesting several times, I have managed to get as low as Cd = 0.36, but with an alarmingly high value for (Ad + Bd = 0.22) , which may well account for the poor mpg that I get.

I don't have the time to go into this further at present, but I will add more, time permitting ;)


edit:
brief explanation of coastdown here --> https://www.quora.com/What-is-coastdown-testing
 
Interesting figures

Sent from my C6603 using Tapatalk
 
Since I've had my accord, I've tried different sized tyres (225/17 and 225&245/19) and tested different pressures and they made sod-all difference to the economy over the same journey at the same speed, and that was the difference between 225's and 245's out the back too.

All the things that I expected to make a difference (including the change in ride height, different types and weights of engine and gearbox oils, replacing all the front suspension components) gave such neglible differences in mpg returned (<0.1mpg), that even when measured over a period of a few weeks I couldn't confirm wether they were effective or not.

The only significant measurable difference (other than driving style) over the last 28k has been where i've bought my diesel, and wether or not it had Redex added or not.

Of course, this may be different for a car with a K24, but for the diesel, it is what it is. Even towing a trailer about with motorcycles on it barely makes a dent.
 
For cars of the same body type and engine, there is always a noticeable range of mpg reported between owners, and I think that this can be found from almost new, indeed possibly from new. I also don't think it's due to driving style quite as much as one would expect.

I also think that this has always been true for all cars. I have tried altering driving style and it has made a difference, but I have never managed to get over 30 mpg, whereas other people report more than 30 mpg for a 7th gen K24 engine. So it can only be due to differences in the engines, or in rolling resistance.

The coefficient of rolling resistance itself, should be far less than 0.1. This is the actual rolling resistance between tyre and surface. But when you do the coastdown test, the bearings and driveshafts and a slightly binding caliper will be included, because you cannot differentiate between those and the actual RR between tyre and surface.

But 0.22 is extremely high. I do know that one of the calipers is slightly binding, and I have noticed that the fuel gauge drops much less when on motorways than when driving round town. I suspect that the reason could be due to brakes being used more round town, and so on motorway, the pad that is sticking will wear off quickly, plus less use, will result in higher mpg. But it can also be said that an engine that is cruising is more efficient.

In order to find out which effect it is due to, i had the idea that we could all do the tests and make comparisons of the overall RR. For the coastdown test, engine type is not important, and as long as the body types are identical and tyres brand and size similar, the differences on overall RR would demonstrate that there are issues present in either one (or more) calipers, or one (or more) bearings, or one (or both) driveshafts.

However, it is very difficult to do the test, the formulas are extremely sensitive to the times measured in the coastdown. A change of 0.5 seconds in the formulas affects the Cd and the (Ad + Bd) quite significantly.

I believe that this might be a significant project for a forum, but my late father's house (where I found the book) needs a lot of work, so my own time is going to be limited in the coming couple of months :(
 
This is one of big problems with the way the correction factors for each individual car are derived for use in the calculation of the economy cycles on the dyno.

In the real world they're skewing the result one way by generating extra drag, so that a greater correction factor can be used, yet driving the standard cycles on the rolling road they're pulling back pads in the calipers and lowering the resistance.

It's how you get Mercs and Vauxhalls claiming 40-odd% more economy than they truly achieve.
 
Post up the method, and I'll give it a go in my diesel wagon with its big boots for comparison when I get the opportunity.
 
Goodluckmonkey said:
Post up the method, and I'll give it a go in my diesel wagon with its big boots for comparison when I get the opportunity.
ok, here goes (and note that I am using a Cd of 0.34 as the "target" result for CM1/CM2/CN2 and Cd of 0.30 as the "target" for CL7/CL9/CN1)

You will need a precise digital readout of vehicle speed, I used something similar to a "road angel", but a Garmin or similar will do (GPS does have a lag, but the lag is not significant at the low speeds and low deceleration used).

You do two coastdowns. I used 40mph to 30mph, and 20mph to 10mph, but with hindsight 20mph to 15mph is better than 20mph to 10mph (it has to do with the segments in the coastdown curve).

You'll need to be very accurate with the stopwatch too (I used an old Nokia 6610i , because it has buttons LOL).

You will need to use a straight, flat surface in dry calm conditions. The coastdown from 40 to 30 takes about 0.2 miles in distance, the coastdown from 20 to 10 (or 20 to 15) takes less distance. But use the same surface for each.

You will also have to be quite accurate with the all-up weight. I think that the ranges of curb weights in the tables in #1 are from car completely empty and no fluids, to car completely empty with full fluids. I think that a full tank of fuel is as much as 50 kg, so if it's not full, take that it into account. Also a spare wheel (if you have one) . Also the weight(s) of anyone in the car.

The data I need is
1. the times measured and speeds used for the upper coastdown
2. the times measured and speeds used for the lower coastdown
3. the all-up weight of the vehicle
4. the vehicle body type (CM1/CM2/CN2 etc) so that I know if the data is giving a Cd close to quoted values.

Note that you need to do several coastdowns in each direction. You will find that one direction is more than the opposite direction, because even if a surface looks flat, it may be on a slight incline (optical illusion). If there is a significant difference in times in each direction, the surface isn't as flat as it looks. Also, if you get significant differences between runs (as I did initially) your technique is suspect.

No hurry, as of this weekend coming, I will be using a Vodafone data link in the next few weeks (no broadband, and poor mobile coverage at the house). But I have a "watch" on the thread so I will see any email notifications etc. But I'm on my home broadband until then.
 
Cool. I'll have a go at some point.

The car's only partly reassembled on the drive after an exhaust manifold replacement, so.i doubt it'll be on the road again till the weekend anyway.
 
After a bit of faffing about in the spreadsheet and some fudging of my own data, I've managed to get Cd = 0.34 and (Ad+Bd) = 0.026 :eek: ...those values are extremely good.

The formulas are most "sensitive" to the values for the lower coastdown, so if I have time, I may redo coastdowns yet again, but with a passenger using the stopwatch and taking notes of the times and direction of each run, while I use cruise control and release it at a marker point. It seems that it is extremely critical to get it right, but I think it can be done, and it could be a very useful test on overall RR. If those new values are correct, then the reason that the mpg is poor on my car is down to the engine, and it's always been like that since I've had the car (12.5k miles in January 2006).


btw, I did a heavy edit of #1 and removed the link to the "excellent website" but forgot to add it at the end :blush:

Here is the link --> http://www.automobile-catalog.com/
2004 CL9 --> http://www.automobile-catalog.com/car/2004/1140980/honda_accord_2_4_i-vtec_executive.html
2004 CM2 --> http://www.automobile-catalog.com/car/2004/1141190/honda_accord_tourer_2_4_i-vtec_executive.html

I even checked out an old car that interests me more than Honda's ever will --> http://www.automobile-catalog.com/car/1952/553475/daimler_consort.html

Wherever they are getting the info from, it must have taken a lot of work to put the info into that site, it even has the engine torque/power curves.
 
Goodluckmonkey said:
This is one of big problems with the way the correction factors for each individual car are derived for use in the calculation of the economy cycles on the dyno.
In the real world they're skewing the result one way by generating extra drag, so that a greater correction factor can be used, yet driving the standard cycles on the rolling road they're pulling back pads in the calipers and lowering the resistance.
It's how you get Mercs and Vauxhalls claiming 40-odd% more economy than they truly achieve.
I think there are some other powers at work when it comes to the Mercedes economy figures.
 
freddofrog said:
After a bit of faffing about in the spreadsheet and some fudging of my own data, I've managed to get Cd = 0.34 and (Ad+Bd) = 0.026 :eek: ...those values are extremely good.

The formulas are most "sensitive" to the values for the lower coastdown, so if I have time, I may redo coastdowns yet again, but with a passenger using the stopwatch and taking notes of the times and direction of each run, while I use cruise control and release it at a marker point. It seems that it is extremely critical to get it right, but I think it can be done, and it could be a very useful test on overall RR. If those new values are correct, then the reason that the mpg is poor on my car is down to the engine, and it's always been like that since I've had the car (12.5k miles in January 2006).
Anyone that has owned a Firestorm will confirm the engine is all that determines economy, it will return 30-34 mpg no matter how carefully/carelessly one uses the throttle.

The 'flat' road specified in the article is of interest though,asphalt will always have its own resistance owing to the way tyre rubber and tread interacts with the particular grain on the surface, built into the specification by MacAdam and improved over the years to improve braking distances among other things (skid prevention).
Additionally a camber free road (driving down the white line) will help with the numbers, as will adjusting the toe in on the wheels, both removing bearing drag via the cage sides.

Lowering the car, fitting a silk skirt that just misses the ground and using a high polish on all paintwork and glass should help with the rolling distances.
 
Channel Hopper said:
Anyone that has owned a Firestorm will confirm the engine is all that determines economy, it will return 30-34 mpg no matter how carefully/carelessly one uses the throttle.
We're taking passenger vehicles so not really relevant to this thread, particularly as passenger vehicles usually have 4 wheels, 2 of which steer, and the vehicle is usually an enclosed shape

Channel Hopper said:
The 'flat' road specified in the article is of interest though,
The link to the "article" is only there to serve as an explanation of the coastdown method, I'm referencing a chapter on the subject out of a book on automotive engineering

Channel Hopper said:
asphalt will always have its own resistance owing to the way tyre rubber and tread interacts with the particular grain on the surface, built into the specification by MacAdam and improved over the years to improve braking distances among other things (skid prevention).
ok, arrange these in order of coefficient of rolling resistance
  • Pneumatic tyre on concrete asphalt
  • Pneumatic tyre on earth
  • Pneumatic tyre on farmland
  • Pneumatic tyre on large sett pavement
  • Pneumatic tyre on rolled coarse gravel
  • Pneumatic tyre on small sett pavement
  • Pneumatic tyre on tarmacadam
  • Railroad wheel on rail


Channel Hopper said:
driving down the white line
do you alert all media before you set out :lol:

Channel Hopper said:
Lowering the car, fitting a silk skirt that just misses the ground and using a high polish on all paintwork and glass should help with the rolling distances.
How so ?
 
You mentioned the poor mpg was down to the Accord engine, and If nothing at all affects the economy, or not of another Honda that has multiple cd variation then all I've done is reaffirm you are not alone.

You left cobblestones off the list.

Lowering any vehicle will remove a portion of the total frontal area (as will chopping two inches off the pillars, the Scirocco Mk1 top speed was higher than the same spec Golf)

Cloth skirting - Used to diffuse the limit layer, installed on close fitting panels around wheels during speed runs in the early 20's, more efficient than perforation or ***isted venting at moderate velocities.
 
Channel Hopper said:
You mentioned the poor mpg was down to the Accord engine, and If nothing at all affects the economy, or not of another Honda that has multiple cd variation then all I've done is reaffirm you are not alone.
The fact that Matt understood, shows to me that my explanation was clear, which is that the exercise is to find out whether the overall-RR for one's car is high or not. For a given car body type (e.g. CM1/CM2/CN2) the Cd is the same, so if 2 people with car of the same body type make coastdown tests and the times for their coastdown tests give the correct Cd but very different RR, then that starts to point to issues to do with RR.

But it is important to know what the Cd value should be in the first place, because that is the only part that will be constant. Thus the thread ***le is based on having found sources online for the Cd of the vehicles, which one needs to know before entering test results into the formulas that I have (because if the formulas produce incorrect Cd from one set of results, then that indicates that any RR from that set of results is also incorrect).

Basically, this thread is about three things
1. finding a published Cd for one's car, which seems to be available online along with virtually most other specs for one's car
2. how to perform coastdown tests to see if the overall-RR for one's car is high , but I'm not going to give out the formulas for obvious reasons
3. it would be useful to compare overall-RR between two cars of the same (unmodified) body-type, if they are similar but one car has better mpg than the other, then the issue has to be due to the engine


Channel Hopper said:
You left cobblestones off the list.
the list that I provided has plenty of variations in coefficient of RR for various type of surface regularly encountered in this country



Channel Hopper said:
Lowering any vehicle will remove a portion of the total frontal area (as will chopping two inches off the pillars, the Scirocco Mk1 top speed was higher than the same spec Golf)
Terminologies - "lowered" usually means reducing the ground clearance, whereas reducing the height between car roof and floor is known as "chopped".

Total frontal area does not include the air under the car, because there is air-flow under the car. Thus for a given vehicle, lowering the car will produce higher Cd because of the "ground effect", whereas chopping the car will produce lower Cd. Note also that one of the main reasons for a plastic undertray is to reduce turbulence under the car, which reduces Cd.

But note that reasons and explanations for Cd are not relevant to this thread.



Channel Hopper said:
Cloth skirting - Used to diffuse the limit layer, installed on close fitting panels around wheels during speed runs in the early 20's, more efficient than perforation or ***isted venting at moderate velocities.
I'm not sure that that would be called "cloth skirting", more like "cloth veneer". But now you mention enclosing the wheels, yes that can be effective on the rear wheels, similar to fitting an undertray.

I wondered if you meant something like this --> https://youtu.be/Rthl0iKG8o8

But either way, my intention for the thread is not to discuss ways of modifying Cd, it's to proffer the notion that there might be a way to find out experimentally if one's car has poor RR. It is the experimental procedure that is important to understand, because the formulas seem to be extremely sensitive to the times for each portion of the coastdown curve, as well as published data on the vehicle.
 
Matt also mentioned that towing a trailer, with extra wheels and additional drag, made no difference to his fuel consumption. My gut feeling is that the run in period (coaxing the piston rings and bores to work in unison) and the way the suspension and tyres interact are the only things that make a difference with vehicles exceeding a tonne on a level road. Again as Matt will concur, with bikes the better the suspension, the lower the MPG since the tyres make positive contact more of the time.

Turbulence under the car is reduced by improving the desgn in the first place (Citroen used covers on strategic parts on the GS to additionally reduce road noise at speed).

Air dams create their own problems which is why they are usually put on cars with enough engine power to overcome the disadvantages.

Anyhow I'm willing to give rolling resistance checks a go (I have a Garmin Nuvi somewhere) but will have to find a level road (and one I can drive down the centre of)

How are you going to take into account the variations of tyre size, roof rack/rails on individual cars, or if one has left road dirt everywhere (like mine) or polished it to be as slippery as a hagfish ?

http://www.bbc.co.uk/news/world-us-canada-40605743
 
Channel Hopper said:
Matt also mentioned that towing a trailer, with extra wheels and additional drag, made no difference to his fuel consumption. My gut feeling is that the run in period (coaxing the piston rings and bores to work in unison) and the way the suspension and tyres interact are the only things that make a difference with vehicles exceeding a tonne on a level road. Again as Matt will concur, with bikes the better the suspension, the lower the MPG since the tyres make positive contact more of the time.

Turbulence under the car is reduced by improving the desgn in the first place (Citroen used covers on strategic parts on the GS to additionally reduce road noise at speed).

Air dams create their own problems which is why they are usually put on cars with enough engine power to overcome the disadvantages.

Anyhow I'm willing to give rolling resistance checks a go (I have a Garmin Nuvi somewhere) but will have to find a level road (and one I can drive down the centre of)

How are you going to take into account the variations of tyre size, roof rack/rails on individual cars, or if one has left road dirt everywhere (like mine) or polished it to be as slippery as a hagfish ?

http://www.bbc.co.uk/news/world-us-canada-40605743

freddofrog said:
For cars of the same body type and engine, there is always a noticeable range of mpg reported between owners, and I think that this can be found from almost new, indeed possibly from new. I also don't think it's due to driving style quite as much as one would expect.

I also think that this has always been true for all cars. I have tried altering driving style and it has made a difference, but I have never managed to get over 30 mpg, whereas other people report more than 30 mpg for a 7th gen K24 engine. So it can only be due to differences in the engines, or in rolling resistance.

The coefficient of rolling resistance itself, should be far less than 0.1. This is the actual rolling resistance between tyre and surface. But when you do the coastdown test, the bearings and driveshafts and a slightly binding caliper will be included, because you cannot differentiate between those and the actual RR between tyre and surface.


In order to find out which effect it is due to, i had the idea that we could all do the tests and make comparisons of the overall RR. For the coastdown test, engine type is not important, and as long as the body types are identical and tyres brand and size similar, the differences on overall RR would demonstrate that there are issues present in either one (or more) calipers, or one (or more) bearings, or one (or both) driveshafts.

However, it is very difficult to do the test, the formulas are extremely sensitive to the times measured in the coastdown. A change of 0.5 seconds in the formulas affects the Cd and the (Ad + Bd) quite significantly.
 
Noted, but why repeat. If enough people sign up to this then you need to create a(nother) spreadsheed detailing wheel/tyre sizes, roof-rail or not, type of road surface etc.

A two way (possibly best of four) run will be required to remove the possibility of the invisible incline affecting results.

All of this down the centre line............................
 
Channel Hopper said:
you need to create a(nother) spreadsheed detailing wheel/tyre sizes, roof-rail or not, type of road surface etc.
you can do that

Channel Hopper said:
A two way (possibly best of four) run will be required to remove the possibility of the invisible incline affecting results.
can't cherry-pick in science, one's driving/timing technique has to be able to produce consistency, consistency is actually very difficult


Channel Hopper said:
All of this down the centre line............................
you can do that, but alert all media before you try anything like this on a public road
 
Top