23 December 2012

"Its been a while..."

Just as this blog was starting to pick up steam I went on hiatus. -Sorry for that. Here are some snap shots of the stories I owe you:

Replaced the muffler on Ian's Honda CBR 600.
Installing brake lines and CPP Master Cylinder in Matt's Camaro.
Getting Matt's Camaro back on the road.
Replacing the roof on Project Identity Crisis.
My tour of a Mopar Haven in Missouri.
As you can see I have been busy! To top it all off my server has kicked the bucket. Something about a guy in an American Flag Speedo disagreed with it...


I will split my time between fixing the server and writing about the stories I owe you. Look for things to start picking back up after the holidays!

Very respectfully,
-JimmyJam

26 September 2012

The LEGO Tire Test

   Many don't realize it. But the LEGO group produces more tires than any other manufacturer in the world. In fact in 2011 LEGO made 381 million tires. The next closest competitor was Bridgestone with a mere 190 million tires. The down side of this is the monopoly LEGO has in the LEGO car model world, where as real cars have lots of manufactures competing for sales. This means real tire makers have to provide proof that their product is superior. However I have noticed a major lack of performance data for any or the LEGO tires. Philo has done some comparative traction tests. And lots of people have published weights and measurements. But beyond that how do the tires affect the performance of LEGO Technic and Mindstorms models?

A selection of tires I will use for these tests. I haven't yet tested the tracks yet. There are many questions to be answered about the tires first.


   Granted there aren't too many people racing their creations on the ragged edge. There are some folks out there that build with function and practicality in mind. So I will delve into a multi part examination of how a selection of LEGO tires which are popular with powered creations can impact how they behave.

   Many times in my life I have heard that big tires hurt your mileage (MPG or Liters/100Km), and that better economy can be had with small skinny tires. However most people don’t think about why this is.


            -How does rolling resistance play in to this?
            -Don’t tall bicycle tires go over bumps easier than little roller skate tires?
            -What about tire and wheel weight? -A top can spin for a long time, usually heavier tops spin longer than light tops. Why doesn’t this hold true with tires on a car?

   There are some fairly obvious reasons why passenger cars aren’t driving around with tall tractor sized tires, but if diameter and mass can help a tire keep rolling over bumps why do people put small skinny tires on cars to improve mileage?

   It would be very expensive for me to test this with real tires, and the price of gasoline would almost be prohibitive. But I can scale it down with LEGO tires and electric motors. Although it is kind of embarrassing to consider the money I’ve invested into LEGO tires, the fact is I have them. And the energy needed to test these small tires comes pretty easily thanks to some rechargeable AA batteries. I also have a decent collection of different LEGO motors demonstrating a wide spectrum of speed and torque. So now I can compare different tire combinations with each of the motors and find the most efficient combination (as well as the least efficient).

   Philo did some excellent work analyzing most of the LEGO motors. He basically built a dynamometer, and used a power supply with variable voltage to examine how all of these motors perform at various voltage levels. From his research we know maximum speed, maximum torque, and each motor’s ability to turn electrical power in to power to move things. If you are familiar with performance cars or hot rodding you may be familiar with Engine Dynamometers. These are great tools to calculate how good a motor or engine is at producing power (and in the case of internal combustion engines; tunning them to perfection). But what you often don’t see on these is how parasitic drag from things like transmissions, alternators, power-steering pumps, radiator fans, and air conditioning pumps as well as driveline and axle losses affect these engines (which often need more tuning to make them drivable in the real world). All of this is done on something called a chassis dynamometer. A chassis dyno has rollers under the car’s drive tires which are equipped with sensors and wired to a computer which can translate torque and speed in to horsepower (the equation is HP = FtLbs x RPM / 5252 if you’re wondering).

   I didn’t build a LEGO sized chassis dyno (although that could be possible…) But I can do the next best thing – build a drag strip! When it comes to motors (or engines), power is a product of torque and speed. If you know how much a vehicle weighs, and you know how quickly it can accelerate for a set distance you can calculate the average power. If you break this distance down you can refine the data to see how much power the motor is producing at a given speed. The mix of LEGO motors I have demonstrate a varying range of speed (RPM) and torque. Some motors turn slow with lots of torque, some turn fast with little torque. I want to see how optimizing things like gearing, tire/wheel selection, and chassis setup can maximize power put to the ground. In most cases we are limited to 9 volts with LEGO motors (7.2 if using NiMH or 7.4 with LiPo rechargeable). So the best way to maximize power to the ground is to maximize how efficiently we use those few volts in LEGO models. That is what leads me to this battery of tests (pardon the pun). And to start off I need to know which tires are best for this and how to best gear each motor to work with these tires, this leads me to my first test.

   There are lots of ways to approach this test. Mine is to find the steady state power consumption of driving a simple LEGO vehicle with different tires and motors. I originally planned on making a 10 meter long track. But I didn’t want to work outside. I also didn’t want to drag all of my stuff to a gymnasium and explain what I was doing to every random person that stopped by. Luck would have it that I have enough room in my house to set up a 3 meter track. Thanks to the benefit of averages, I could simply run each combination down this track say three times for a total effective length of 9 meters (pretty close to my original plan) and still get a good chunk of data. Now I have five different motors, and 17 different types of tires for this test. That results in 85 different motor/tire combinations, then I ran each test three times that’s 255 runs. Each run was three meters, so I drove my little LEGO vehicle 765 meters! That’s about 0.48 miles, which isn’t far to walk or drive, but is a long way to chase a little LEGO vehicle for sure.

   The motors I’m using are from fastest to slowest (in terms of output shaft RPM) are:

5292 RC Buggy Motor

9670-1 PF E Motor (from the LEGO Education “Energy Meter” set.)

8883-1 PF M Motor 

PF L Motor (new with the LEGO Rock Crawler)

8882-1 PF XL Motor


The tires are:

Part # Style Size Mass Inertia
2696 Street 13x24 14g 1.391
2857 Balloon 20x30 19g 1.753
44309 Street 43.2x22 ZR 15g 1.424
6579 Balloon 43.2x28 S 17g 1.486
55976 Balloon 56x26 17g 1.851
41897 Street 56x28 ZR 24g 2.192
32019 Street 62.4X20 S 33g 2.775
61480 Balloon 68.7x34 30g 3.136
41893 dull 68.8x36 Hd 36g 3.425
41893 shine 68.8x36 Hs 36g 3.478
44777 PLASTIC 68.8X36 ZRp 18g 2.711
44771 Street 68.8X36 ZR 39g 3.612
2902 Balloon 81.6x15 29g 3.714
45982 Balloon 81.6x38 R 40g 4.288
92912 Street 94.3x38 R 68g 6.207
54120 Balloon 94.8x44 R 58g 5.833
51380 Lg MtrCyc Ft 36g 4.656


Mass of each tire was measured on a simple kitchen scale, I’m sure there is some error here, hence I didn’t bother with decimals. The inertia was determined with a weight drop test. In this test I made a simple setup that I could place on a counter and hold a LEGO axle a little more than one meter above the floor. I then filled a medicine bottle with ballast to make a 100 gram weight. I tied this weight on to a string and wrapped it around the LEGO axle, being careful to not let the string double over its self (just one long wrap). I then placed each wheel on the axle one by one and timed how long it took for the weight to drop to the floor. This was done with a stop watch and is not super accurate but does serve to demonstrate how the larger wheels are harder to accelerate than smaller wheels. This would apply to real world mileage if you did a lot of stop-and-go driving. I may perform a test later to see how much energy it takes to accelerate each tire up to a given RPM.

Armed with this inertia data I decided to keep first test slow, literally. I chose a gear ratio that was simple (for the sake of consistency) but also was a decent compromise for the performance of all of the motors. My three big concerns were: not going so fast with the RC motor as to spin the tires; not to bog the soft hitting PF E motor; and not to take a whole day with the strong, but slow PF XL motor. I found a 27:1 ratio easy to build and a happy medium for all of the motors. With the RC motor and the largest diameter tires the vehicle covered the three meter track in about 12 seconds, with the PF-XL motor and the smallest tires it took a little over two minutes and thirty seconds!

This test was very time intensive, but necessary to pave the way for follow on tests. With this test I was able to see how rolling resistance of each of the tires impacted the efficiency of the test vehicle. Below are videos of two of the runs I did, the only difference between the two is the motor used. The first video is with the RC Buggy Motor (fastest of the group), the second is with the PF XL Motor (slowest of the group). The difference in speed is very obvious.






Lets take a closer look at the tires:

2696 - 13 x 24
 2696 is a solid rubber tire made for "Model Team" vehicles and saw some use in several of the Technic Universal building sets as well. It is a hard compound and can hold a lot of weight. It is the lightest tire in the test and has the lowest moment of inertia (MoI).






2857 - 20 x 30
 2857 Is also an older tire, and is also solid. It is a softer compound than 2696 and had larger voids between the tread. the wheel is a smooth cylinder and can slide out or spin inside of the tire.










44309 - 43.2 x 22 ZR
 44309 is one of the smaller semi-pneumatic (s.p.) tires, and is what you will find in an NXT 2.0 set. The wheel is shared with the 55976 tire and can be used with rubber tread (caterpillar tracks). It's the lightest s.p. tire in the test and has very low moment of inertia (MoI).





 



6579 - 43.2 x 28 S
 6579 Is an old balloon tire. I think the four I have are from a 2000 or 2001 set.









55976 - 56 x 26
 55976 is a very common tire and is also found in the NXT 1.0 sets.









41897 - 56 x 28 ZR
41897 has the same diameter as 55976 but the larger wheel moves the mass closer to the outer edge, this tire has slightly more weight and MoI than 55976








32019 - 62.4 x 20 S
 32019 is often used for it's realistic look and heavy load capacity. The wheel has two circumferential ribs that act to stabilize the tall sidewalls. The wheel and tire have a lot of mass for their size, but the MoI is still less than the larger diameter tires.








61480 - 68.7 x 34
 61480 is very common due to Technic's push for tractors and construction equipment a few years ago. It uses the familiar staggered void tread pattern but seems to have fairly high rolling resistance compared to other tire near it's size.







41893 - 68.8 x 36 H
41893 Dull and Shiny - I have two sets of these wheels. The dull rubber ones (pictures with gray wheels) came with my RC buggy and were made for play outside. I'm not sure where I got the shiny ones, but it has the more common tire compound. The difference in weight was negligible. The dull tires had slight but consistently lower MoI. I haven't compared the traction values of these tires yet.

 In a strange way the slower motors seemed to prefer the shiny version, while the faster motors moved more efficiently with the dull tires. I hope to spend some more time comparing these two.













44771 - 68.8 x 36 ZR & 44777 - 68.8 x 36 ZR Hard Plastic Tire
44771 and 44777 are the same diameter, and made to look similar. While 44771 is a typical s.p. tire, 44777 is a simple and light plastic one-piece tire and wheel. The later is intended for "stunt driving" with remote controlled models.





44777 is only 18 grams compared to 44771's 36 grams, and the MoI reflects this by being almost 30% lower. However the slippery, non-conforming nature of the plastic stunt wheel makes it slightly harder to roll, as reflected in the Joules per meter charts.







2902 - 81.6 x 15
 2902 Made for motorcycle style creations, it is a large diameter soft compound tire with a narrow section width. Like 32019 the wheel has several circumferential ribs to stabilize the tire with heavy loads. The narrow width keeps the mass down relative to other large diameter tires, however the design (specifically those ribs) give it a pretty high MoI.





45982 - 81.6 x 38 R
 45982 Getting in to the larger tires, this one is common with larger Technic creations. It is a tall profile tire (relatively tall sidewall compared to wheel diameter) and is soft. It seems to give under load easier than almost any of the other tires in the test. It uses the same wheel as the 41897 & 41893 tires. It is one gram lighter than the 44771 but has higher MoI due to it's larger diameter.






92912 - 94.3 x 38 R
92912 Was created for the LEGO Unimog and is the heaviest tire in this group. It's tread patter is similar to 32019's, simply made larger. This tire shares the same wheel as the 44771 and 54120. It has the highest MoI, even though it's diameter and width are less than 54120. It is a very sturdy tire and appears to hold more weight than any other s.p. tire in the group. It may be well suited for a heavy creation but takes the most power to accelerate.



54120 - 94.8 x 44 R
54120 Is the overall largest tire in the group even though it is 10 grams lighter than 92912 (58 grams vs 68 grams) as expected it's MoI is high, but its large diameter lends it well to creations that need to cover a lot of ground. Although stopping and going consumes a lot of power.






51380 - "Tire Large Motorcycle Front"
51380 Matches 54120's 94.8 mm height yet has half of the mass of 92912. This was the tire I expected to have the greatest sustained speed efficiency of the group.










Here's some visual aides of how the tires and motors compared to each other.
The first thing we will look at is the speed of the vehicle with each tire. Remember all motors and tires were run with the same 27:1 gear reduction. Philo recorded the no load shaft speed of the PF M and PF L motors to be very close (405rpm and 390 RPM respectively), this graph reflects that. Also of note is the difference between the PF E motor, and PF L motor. Again this graph shows that E motor is twice as fast as the L (780 rpm vs 390 rpm respectively). This graph doesn't demonstrate any one tire as better than the other, it merely serves to demonstrate how changing tire diameter and nothing else will affect vehicle speed (not counting for impacts on acceleration or wind resistance). But I wanted you to see how motor speeds recorded by Philo are echoed in this test, however with very light loading on the motors their true potential is not yet displayed. Again this is just a slow baseline.

This graph is the most literal equivalence of miles per gallon (km/100L). A Joule in the terms of electricity is one watt for one second. Since the taller tires increased the vehicle speed the motors had to do less revolutions to cover the distance, thus improving economy. This data is from the same test runs as the graph above. Ideally what we want to find is a tire that minimizes the Joules used to travel a meter. The very slow gearing used for these tests favors the motors with higher shaft speed, and makes the torquey L and XL motors look like hogs (the truth is they had lots of mechanical power to spare and simply need faster gearing). In this first test the E motor looks like it sips Joules per meter compared to the other motors. One very interesting thing to take away from this graph is how the RC and M motors nearly matched each other perfectly. This is despite the fact that the RC motor is 2.5 times faster than the M motor. My prediction is this scenario is close to ideal for the RC motor, and the M has room for improvement with better matched gears.

The next set of charts display each motors rate of power consumption (Joules per Meter), with each combination of tires. A common thing I noticed through all of the test was how the power draw would usually drop slightly from the first test to the last test. My method of testing to put a set of tires on the vehicle then conduct three runs with each motor, the first run was the 13x24 tire with the RC motor. I ran this test three times, then turned the energy meter off while changing to the E motor. Again I would run the vehicle three times, then shut off the energy meter to change to the third motor and so on. I'm not yet sure why the economy improved, perhaps the motor turns easier with a little warming up. Or maybe the energy meter simply flows more watts when it's cold. My hypothesis leans towards the later possibility since the phenomenon was more pronounced on the more powerful motors. For each of these graphs the vertical bars display the average wattage of each motor/tire combinations during the three 3m test runs, and the purple bar is the average for each combination over the three runs. Lower numbers mean fewer Joules per second, the following info basically shows how much power the motor was consuming to turn the tires.

RC Motor specifics:
The RC motor has an almost wild and un-predictable personality, which is demonstrated by the variations in data show above. For my tests I followed Philo's lead and used the outermost output for testing. At first glance this data may appear to be useless. However when you group similar diameter tires and compare power used the individual tires ease of rolling starts to stand out. Take for example the 56x26 and 56X28 ZR tires. These are the 55976 and 41897 respectively. These two tires are the same diameter yet on this graph the later was using less power to turn. If you compare this to the speed graph for the RC motor you sill see that the vehicle traveled slightly faster with the 41897. Faster speed and less power means greater economy. Try comparing the last two tires in the same fashion

E Motor specifics:
The PF E motor had the overall lowest rate of power power consumption, and was pretty steady with a few exceptions. Notice the 20x30 (2857) and the 81.6x15 (2902), both have soft compounds and aggressive tread with deep voids which has a negative impact on economy. Again compare the last two tires. The 54120 and 51380 have the same overall diameter and produced similar speeds but the 51380 consumed less power to do so. 
M Motor
The PF M motor was pretty agnostic to tires with this gearing with two exceptions. I seemed to draw less power turning the 43.2x22 ZR (44309) and 56x26 (55976) both consumed less than .5 watts per second. Now compare how much faster the vehicle traveled with the 55976 tire; that is a step in the right direction.
L Motor
The PF L motor produced similar speeds to the PF M Motor, yet consumed nearly twice the power. Also this graph demonstrates more clearly the trend to use more power on initial runs then on following runs. This may be the motor I use to evaluate this phenomenon.

XL Motor
The PF XL motor averaged about 60-70% of the PF L Motor's speed with a similar reduction in power consumption. Philo noted just slightly higher efficiency out of the XL than from the L motor (45% vs 42% at 9 volts). Judging by the rate of power consumption the XL motor was operating at near "no load" speeds. Faster gearing will improve efficiency.
In summary this test collected data on how tire diameter can influence vehicle speed, and how each motor responded to the rolling resistance of each respective tire. The next bit of research I do will be more exploratory in nature. I will choose 44309 (43.2 x 22 ZR), 41893 (Shiny version of 68.8 x 36 H), and the 51380 (Large Motorcycle Front), as they seem to have favorable ratios of meters per second vs Joules per meter. For these next tests I will reuse the same 3m track and vehicle chassis. And I will work to develop optimized gearing for each motor, in a sustained speed scenario. As always feel free to leave a comment!

19 September 2012

My new MINUTEBOT Base

Just got a new toy in the mail today, it's called a MinuteBot Base and is the first product from a startup company in Denmark. Here is the link to their site. Here is a short video of my first (a-hem) creation with it. I learned about this on the TechnicBlog when it was just an idea posted on the Kickstarter website. I like to support underdogs and knew I could have plenty of uses for this, so I made a plege. Towards the end it was looking like it wasn't going to get the funding it needed, but in the last few days someone pledged a nice large sum that put the project comfortably in the green. They plan on producing a few more things to help bridge the gap between the fun and playability of LEGO Mindstorms robotics with the precision and accuracy of industrial robotics.


 I've also made some good progress on my LEGO tire test. Here's a picture of the vehicle -er test rig. that I will use to compare the tires. It has places to carry extra weight, and a quickchange rear axle and motor set-up. I plan on comparing several common sizes of LEGO tires, gear train designs and motors. It is all based around a new LEGO Education item for LEGO Mindstorms. It's called the LEGO Energy Meter and can measure voltage, amps, and watts coming in and going out. When you use it with the Mindstorms NXT intelligent brick you can log the data to make comparisons, I think you see where I'm going with this.

The smaller wheel in the middle is for measuring distance. You can see some of the tires I will test in the corners of the photo.


Here is a little video made by LEGO Education demonstrating the LEGO Energy Meter.

15 September 2012

Changing spark plugs on our 2005 Ford Freestyle



Replacing the sparkplugs, and Transaxle Roll Restrictor mount on a Ford Freestyle

Why change the plugs? Mileage was down; start-up crank times were starting to get a little longer than when the car was new. Plus reading the plugs can give you a good indication of how your engine is running.

Why change the Transaxle Roll Restrictor? Apparently it’s pretty common for the rubber bushing on these to wear-out. This mount is on the top of the engine and is easy to replace if you are already doing spark plugs.

Time: 6 hours

Tools needed:
-3/8 inch ratchet with 10mm, 13mm, 19mm, and 3/4 inch sparkplug sockets and a few extensions.
-1/4 inch ratchet with 8mm socket and T-25 torx bit and a few extensions.
-8mm, 1 1/16 inch combination wrenches, as well as a large adjustable or 24mm combination wrench.
-Needle-nose and standard slip joint pliers
-1/4 inch and 1/8 inch tip slotted screwdrivers
-sparkplug gap tool (spec is .052-.056 –I got a gauge that has .054)
- Catch-pan for draining coolant (I used and old litter box)
Recommended:
-3/8 inch air ratchet and blow-off gun
-1/4 inch driver handle
-High-temp anti seize compound
-shop vac

The shop vac and large crescent wrench are missing from this picture.


Parts needed:
-Transaxle Roll Restrictor -$260!!!
-Upper intake manifold gaskets
-I STRONGLY recommend lower intake gaskets as well.
-Spark plugs (recommend platinum or double platinum, when in doubt ask the guys that work on cars at a Ford dealership, not just the guys that sell parts)

First step, make sure you have everything you need, including a second vehicle to make a run to the parts store. Check?

Okay, this can be done in any driveway or garage, and with the exception of a few large wrenches it can be done with the most basic of hand tools. The first big step is removing the big cross brace the Transaxle Roll Restrictor (TRR from now on) mounts to. It needs to come out to get the upper intake manifold off of the engine. It’s not too crazy of a thing to do; there is one bolt where the brace connects to the TRR, then four bolts that connect the brace to the strut towers. These bigger bolts are 24mm, I have a socket that fits them but there isn’t enough clearance for this particular socket and ratchet (from my ¾ inch drive collection). So I just used my large crescent wrench to pull these bolts, they aren’t supper tight and are deceptive to look at, the bolt head seems much larger than it needs to be. Once you pull the five bolts you can remove the brace and set it aside. I make a habit of placing the bolts back in the holes they came from so I don’t lose track of them, just turning them in a few threads will suffice.

The instructions my wife found online provided minimalist instructions for the things that need to be done to get the intake out. If you have been at these kinds of projects for a while you may recall changing the sparkplugs on your old truck or car never took 6 hours or involved removing an intake. True, on an older V8 rear wheel drive car I could often change all of the plugs in about ten minutes. But you never had to remove the intake on those to get to the plugs. It’s a packaging versus physics thing we are dealing with here. The part at the back of the intake with all of the ribs is called the plenum. This helps the engine perform well at all speeds. The smoother pipes between the plenum and the engine its self are the runners. There are equations that explain how longer runners help provide more torque at lower engine speed (or any speed really), but I’m not going to get into those here. Basically Ford engineers determined this was the most effective way to package the intake manifold into the engine compartment on these cars, and all things considered six hours to remove and replace the intake manifold, sparkplugs (and six ignition coils) as well as a motor mount would be tough to beat on your old truck with a small block. In fact good luck doing all of that and not have to redo anything because of a coolant or vacuum leak afterwards. And I didn’t need a drop of RTV sealant.
Here are the basic steps we found online, they are from a Ford Workshop Manual:
1.  Drain the cooling system. (About 1-1.5 gallons will do, these cars have coolant circulating around the throttle body)
2.  Disconnect the battery ground (NEG) cable. (I ended up removing the whole battery and its tray, you’ll see why in step 11.)
3.  Remove air cleaner outlet pipe.
4.  Remove the 4 transaxle roll restrictor (TRR) cross brace bolts.
5.  Remove the bolt and the TRR.
6.  *Disconnect the evaporative emissions (EVAP) canister purge valve tube from the upper intake manifold and detach the retainer.* (This can be ignored, I traced the hard plastic EVAP tube to the passenger side strut tower where is connects to the purge valve. The connections are snap-to-connect fittings and can be released by pressing in on a rectangular button on the side of the connector. I left the tube attached to the upper intake manifold.)
7.  Disconnect the positive crankcase ventilation (PCV), brake booster and vacuum harness tubes from the upper intake manifold. (This is a little vague; see the pictures for a better explanation. Also the rear vacuum tubes will need to be disconnected but I found it easier to wait until I could bring the intake towards the front of the car and reach them easier.)
8.  Detach the fuel tube retainer from the upper intake manifold (it was easier to unlatch and open the retainer, see the pic)
9.  Disconnect the exhaust gas recirculation (EGR) system module electrical connector and vacuum tubes).
10.  Detach the vacuum tube from the upper intake manifold vacuum tube retainer. (this is a sneaky little guy, it’s one tube wrapped in corrugated conduit and basically jammed into a slot down on the lower driver-side of the upper intake just above the lower intake manifold, I could not get it back in that slot the way I found it.)
11.  Disconnect the EGR system module tube from the EGR system module. (This is the one significant got-cha in the whole project, it takes a big wrench of a specific size and it is really rusted on there. If you round the corners on this you are in trouble, if you get it free you are home-free. Remember this thing is upside down, take a second and consider “lefty = loosey” would be true if you were under the EGR module. You need to spin the wrench from the master cylinder towards the front passenger tire here.)
12.  Disconnect and plug the throttle body and PCV coolant hoses. (This is verbatim from the Ford manual, but if there is no such PCV coolant hose(s) and if you drained the coolant right as per the tips in the pictures you will have little to no coolant in the hoses anyway.)
13.  Disconnect the throttle body electrical connector.
14.  If equipped remove and discard the four upper-to-lower intake manifold screws. (This is what the T-25 bit is for, they are like wood screws that held the two halves of the intake together for final assembly, and serve no structural or meaningful purpose now. You really can toss them.
15. Remove the 8 bolts and the upper intake manifold. *remove and discard the gaskets (at this point there is nothing but gravity and wishes keeping the lower intake manifold in place. Which means it will get bumped around as you wrestle the upper intake out of the car. Hence you should be prepared to remove and replace the lower intake gaskets as well. Also be sure to have a crevice tool on your shop vac and some means of cleaning the sealing surface, like a sturdy rag or one of those 3M gasket cleaning pads from the auto parts store.)

   So now you can get cracking on those spark plugs or that TRR mount, I did the plugs first, but it doesn’t matter at all. If you’ve ever changed sparkplugs before this will be familiar to you, with the exception that you have to remove the coil pack to get to them, which is just one small bolt. Be sure to vacuum or use compressed air down in the sparkplug well before removing the spark plug. These engines don’t like small rocks in the combustion chambers. I took three coil packs out at a time, you can do them one at a time, just be sure to put them back in the right place (torque coil pack bolts to 53 INCH pounds). Make sure your engine is cool before doing this. I’ve tried swapping plugs on a warm engine with plug wells like this and my socket got lodged in there. Something about heat related expansion… I had to wait for the engine (and socket) to cool before the socket would come back out. Remember to gap the plugs between .052-.056”

   As for the TRR, it is pretty basic stuff too. Four bolts plus a little wiggling and it’s out. It serves to control the engine from rolling backward into the firewall when you floor the gas pedal, and doesn’t hold any weight of the engine at all (if it had things would have been interesting pulling that big cross brace earlier on.) I usually prefer hand tools while working, since I like the control and ability to hear my music, but for this bugger I broke out my air ratchet.

   Putting things back together is the reverse order of disassembly, I’ll provide the highlights:

Intake manifold bolt torque: 89 INCH pounds
   Pattern: 
                             Firewall 
Passenger side  (5) (1) (3) (7)  Driver side
                         (6) (2) (4) (8)
                           Front of car

EGR tube torque: 30 foot pounds
TRR Cross Brace (to strut tower bolts): 41 foot pounds
Cross Brace to TRR bolt: 35 foot pounds

   Now for the pictures!

I jumped around a little, and pulled the intake off first, two latches on the filter box and one hose clamp at the throttle body. Oh and the connector for the mass air flow (MAF) sensor and PCV tube (seen laying across the battery).

The bolts aren't as large as you would think. Ford recommends cleaning and applying new (blue) thread lock.

There's the crescent wrench...

Now for the tougher jobs...

The white connector is the EVAP tube that I decided not to disconnect later on.

I recommend pulling the back coolant hose first (come us between the throttle body and the brake fluid reservoir.

My catch pan missed a little, but some towels made up for it. I'm not impressed with the radiator drain on this car.

The drain needed the 19mm socket to get loose, it drains through the center. Use the ratchet to crack it loose then finish opening by hand to spare your tools a coolant bath.

Many of the connectors have a red security tab, you have to pull it down to unlatch the connector.

Here is the snap-to-connect fitting on the brake booster, here I'm pressing the release button, notice the latch inside has a small gap under them.

Releasing the button allows the latch to return, compare this picture to the previous one.

the large vacuum hose on the upper intake has a break-over latch, after a little examination you will figure them out.

This is the other end of the hose that went to the top-rear tube near the throttle body, it had some coolant in it but nothing major. I only removed it to improve access to the EGR tube.

Here's a close up of the TRR and its bad rubber insert. If you have a Freestyle, Mondeo or Five Hundred that is a few years old your TRR may look like this.

Here is the fuel line retainer, you can try to disconnect from the upper intake manifold, but I chose a different route.

You can insert a small screwdriver in the side of the latch on this, then just pop the fuel line out of the retainer, this was easier on my car.

This is the high pressure fuel connection to the fuel rail. I don't recommend messing with it, there is no reason to for this job.

The tube with my finger on it is a breather for the transaxle, it is not supposed to be connected to anything. Its plastic clip just keeps it in place and needs to be detached from the hose on the throttle body, don't for get it when you put things back together.

The battery tray has four bolts holding it down, and a devious little clip.

These three harness connectors need to be pried off the tray, be gentle with the tapped on connectors.

Here's that devious little clip. I didn't pry it out, instead I just grabbed the battery tray and yanked it upward.
After getting the tray out the harnesses kept falling down in my work space. I used a bungee to pull it back.

I hooked it on a wheel spoke.


The hex above the brown tube is the EGR connection.

Be sure to spin the wrench from right to left.

Here the flare nut is slid down the EGR tube. This is the trickiest part of the whole project.

The electrical connection on the back of the EGR is kind of hard to reach, but pressing on the top will release the latch, then you can separate it.

I pulled the intake forward to access the vacuum tubes.


With the intake resting on the front valvecover and radiator you can use both hands to pry the top of each vacuum tube connection with a screwdriver and the bottom with your fingers.

Here's a bottom view of the back of the intake manifold. The EGR is to the right.

Another shot with the EGR and throttle body.

A view down the lower intake manifold.

Press the top tab to unplug the coils.

These plugs look decent for their age, they are seven years old and have about 90,000 miles on them, the color is good, but the gap had opened up to over .070" due to erosion.

After swapping the plugs I started swapping the TRR. I don't have a "cheater pipe" but this is how I get extra leverage when I need it.

Side by side of old TTR and new TTR. There are very subtle differences (the voids are smaller on the new one).

Here's a glimpse of all of the junk that gets between the lower intake and the cylinder heads, This is why I recommend getting new lower intake gaskets. Both of these surfaces need to be cleaned before reassembly.

Getting the intake back into place with the bolts hanging down is a pain, try a short strip of masking tape or electrical tape between the bolts and the intake. Make them so you can pull them out the front after the intake is in-place.

An example of the tape holding the bolts (sleeves really) up to help you out.

This 1/4 inch driver handle isn't necessary but I find it handy. It has a drive socket on the top to attach a ratchet handle.


In summary don't forget to hook everything back up. Including putting that transaxle breather tube back up by the coolant hose near the throttle body, and topping off the coolant. It took a few warm up cycles to work all of the air out. If you have any questions, please leave it in a comment!