Body work with a slight change of direction

It's been a long time since an update to the blog.  Most of Christmas was spent putting up a ceiling in my shop so I could insulate the attic.  Michigan winter was starting to show it's hand and I was getting cold!  Work also found a big way to get in the way of my free time.  I took a new position at work, which is really exciting, but it called for about four weeks away from home to do some testing.  This pushed a lot of other weeknight things into my weekend so I barely made progress in the shop for awhile.

So, after throwing around all of those ideas related to fender flares, I made my decision.  No fender flares!

I know!?  How can I tease all (6) of my readers with grand plans and then not do anything cool/new/out there?  Because I'm going to do some minor tubbing work to the inside of the wheelhouse - not the outside.

My good friend Shaun over at Street or Track has a nice write-up on the process he used.  Essentially, he cut out the inner wheel house and moved it over to the frame rail.  He then filled the gap with sheetmetal.  Great idea, but the problem is the frame rail location.  It's really close to the inner wheel house in the rear, so that's the limiting factor.  In the front he got an impressive 3" of extra width.  In the back - only a single inch wider.  Bummer.  While Shaun's happy how it turned out, there's always that thought that it could be more.

With this result in mind, I looked around and found the Detroit Speed Deep Tubs.  They solve the close frame rail issue by simply moving the frame rail over!  The do this from about the middle of the stock axle bump stop to the back of the wheel house.  They provide a new rail that just goes straight back and then has a quick jog back over to meet up with the rear portion of the stock rail that has the leaf spring shackle mount.

The rail in my kit sitting on top of my driver's side frame rail

The DSE tub installed as seen in a YouTube install video I found.

Here's a video of the install that I found.  With their rail as provided you get an even 2" of extra width across the full length of the tub.  You also have the option of narrowing their rail 1" and then you get 3" wider.  The 2" tub lets you run a 295 tire, the 3" tub lets you run a 315 tire in a '65-6 chassis.  Later years start out with wider rear wheel houses, so they can fit ridiculous tire.

These are the parts that come in the DSE kit.

I'm not using the leaf springs anymore since I'm going to be installing the Street or Track 3-link, but I like the idea of the bushing sleeve being there for the long term.  Who knows, it might be handy to throw some leafs and an axle in there later to move the car around.

I think you could actually do a DIY mini tub with a combination of Shaun's process and the DSE frame rail move if you just made your own frame rails from the bump stop straight all the way back or even make your own jog in the rail.  You could then cut out your stock inner wheel house, move it over to the frame rail and then fill the gap with sheet metal like Shaun did.  I liked that the DSE kit comes with all new metal, so it's another opportunity to get rids of rusty metal from the car.  Plus, time in the shop is getting more limited for me so the kit was the way for me to go on this one.

So what about that 3 link?

An important point is that this new frame rail section from DSE is just about precisely where the Watts link mount for the Street or Track 3 link will be mounting.  This isn't their fault.  I would never expect them to design around a competitor's product.  The great thing is that I'm going to get my Watts link frame without powder coating and without complete welding so I can modify it to fit the DSE rail.  Below are a couple of pictures from a quick test session with a prototype Watts link frame fit up against the stock frame rail so we could see how much of an issue this will be.

With the front of the frame rails being close to the same place, but the rear of the rail being the big change (inward), the upper shock mount won't move much.  I can then move the U-channel of the Watts frame over to fit on the DSE frame and keep the center section of the Watts frame in the right place.

So that's the plan, but what's next?  Cutting out lots of spot welds!  None of this is earth shattering news or a late breaking how-to, so pictures can do most of the talking.

First, the trunk floor came out.

Then I cut out the driver's side rear quarter

Then the rear tail light panel came off

Then the passenger rear quarter came off. 

In the picture above you can also see that I cut out the panel between the rear glass and the trunk lid.  There was some rust in the bottom corners of the window and the underside of that panel that I would never get to any other way.  Out with the old and in with the new!

Then the driver's side wheel house - all as one piece

Then the passenger wheel house.  Notice that you can see the chair against the wall all the way through the car!

I'm going to use the 'full' new rear quarter panel, so this has the piece that goes up under the roof panel in the C-pillar area.  This was actually a bit tricky to get apart and I've never seen detailed pictures of this online anywhere.

These is a box section inside the C-pillar that comes down out of the roof and sits on top of the outer wheel house.  This means that the spot welds between the roof and the quarter panel actually go all the way down to the box section.  In the rear of each side, there's actually another piece in the mix between the quarter and the box section with hidden spot welds!  When I didn't know about this taking the first quarter off I got a bit aggressive with the chisel and I have some work to do to flatten out that piece under the roof panel before this all goes back together.

The piece that surprised me is on the rear side of the box section coming down out of the C-pillar

This is the piece that surprised.  Up inside, above the row of spot welds, I have to flatten it back out somehow.

Another small snag is that when I was heating up the lead joint between the roof and quarter I found this small pinhole of rust under the lead that was forming from the inside out.  I'm working out a plan to get at this rust between the layers.  I may be giving chelation a try injecting the solution from the inside through these layers.  It's an interesting process that I heard about from fellow Detroit area Mustanger, Sven.  You can check out his blog and his use of the process at this link.

So once all of this metal was out I spent an entire weekend with wire wheels and grinders and it barely looks like I got anything done.  It was thankless work, but I'm at least close to fitting the new sheet metal.  I did also get the trunk floor screwed in place - at least a good initial starting point.

In fact, Shaun came over Monday night and we fit up the passenger rear quarter and then the (Ford stamping) rear tail light panel!

Process and Plan

I spent a good bit of time figuring out what order I should do things at this point.

My first decision was when I should do the DSE frame rail mods.  I decided on doing the DSE work absolutely last.  If I cut the frame rails off at this point I'm going to lose the only reference points that I have left - XYZ the location of the ends of the frame rails.  I'm actually sleeping better now that I've made this call.  I had considered fixturing that location and doing the frame rail mods first.  I'm not sure it would be close enough.

Next I was trying to decide what goes first, the outer wheel house or the rear quarter panel.  I've decided on fitting the rear quarter panels first.  This way I'll be happy with their location and how 'straight' the car is as a whole.  I can set the trunk lid on there and check gaps, get the rear valance gaps and fitment, door fitment, etc.  Then, once I pull the quarters off and I've fit the outer wheel houses to a decent point I can set the quarter on and see how much the wheel house is pushing the quarter around compared to my original marks for the quarter fitment.  I was afraid that if I set the wheel houses first I would be fighting another variable to get the quarters to fit how I want.

Of course, the welder will be far far away from the car until I have every panel (except the DSE parts) in place and screwed, cleko'd, and/or clamped in place and I'm happy with the overall fitment of the sheet metal.

I can then take it all apart after tons of marking and work from the inside out.  There are lots of welds that need to be done before the quarters are in place.  Also, while it's apart for that last time I can paint the inside areas that I'll never get access to again.

I think with the process above the body will be in the normal, fully welded, condition to start the DSE deep tub install.  I'm just going to leave the trunk floors screwed to the top of the frame rails even if the floors are welded everywhere else.  This way I can make a singe vertical cut on the frame rail (near the stock bumpstop location), pull some screws out and then the frame rails will drop out off the car.  I have new fresh frame rails to put in the back so I can eliminate even some more minor rust that was lurking in there.

2012: Rear Fender Flare Concept

Back on the unibody, I just stripped out all of the rear suspension to make room for the 3-link coming soon from Street or Track.  While I was under there I decided that it was finally time to attack some of the rust that has been there, but never corrected properly.  I don't even wash this car because this area leaks and is just an all around mess.

Here are the major trouble spots:

The driver's side drop off - looking at the inside of the quarter panel

The rear lower tip of the driver's side outer wheelhouse

The passenger side drop off - looking at the inside of the quarter panel.
There is a big bondo patch still on the bottom

To get after things first I pulled my fuel tank, pump and filters out.  That went easy enough, so next I removed the trunk floors.  Back around 1996 I had installed new trunk floors, but at the time I didn't have access to a welder so we leaned on a friend of my dad's who had a rivet gun and a limitless supply of pop rivets.  I will say that the panels fit really well and it was surprisingly solid.  I won't say it was aircraft quality, but not bad for a 17 year old's first major panel replacement project.

I started drilling out the rivets, but soon just grabbed the grinding wheel, ground off the heads and then pushed the rivets through with a punch and hammer.  A couple hours later all the panels pulled right out.

The trunk with all of the floor removed

OK, so how to proceed?

My major rule for myself if that if I do something to the car, it should fit within the NASA American Iron rules so ultimately I can have a place to run this car.  In this case, you are allowed to run up to a 275 mm wide tire.  Typically they are run on 9.5" or 10" wide wheels.  My current wheels are 8" wide and I run a 245 mm wide tire and that's about all I can fit in there.  It's time to make some room!

The inside of the wheelhouse is rather easy.  Shaun from Street or Track just did a nice write up about how he moved his inner's inboard all the way to the frame rail.

He gained a ton of room (3") in the front:

And a little bit (1.25") in the back:

I plan to follow his process later, but for now I want to leave the inner wheel houses in place as a reference point for my quarter panel work.

The rear window also has some pinholes in the lower corners, so I want to take a look at that and fix that while I'm at it.  This is all pointing me to the newer style rear quarter panel that is the complete piece that welds in all of the factory spot weld locations.

Here is a comparison of the two quarter panels available.    You can see in the top one (the complete panel) how it continues over the top all the way to the side of the trunk lid and also includes the end cap.  The bottom one in the picture is just the 'skin'.

I picked up both at NPD today to have a look at them and decide which way to go.  So far I like the complete panel, not the skin.  The panel is straighter overall too and it just feels like the right way to go.

Actually things got crazy at NPD and I filled up the van!  I also picked up all new trunk panels, both a 65 and 67 outer wheelhouse (for comparison sake) and in the bottom of the picture you can see a 68 fender flare in bare steel.  We'll get to that piece in a minute!

After a few days of hounding Mustang buddies about what I should do, the consensus is that I should take the lead from a guy on Stangnet who put a 67 outer wheelhouse behind a 65 quarter panel skin and stretched it out to make it work.  He fit a 275 mm tire on a 9.5" wheel.

Now that I had both outer wheel houses in my hands, let's see how they compare.

First, the '65-6 piece.  It measured at 6.375" wide

Next, the '67-8 piece.  It measured 7.875" wide!  1.5" wider, just from the wheel house!

Here is a side by side of his car with the 'stretch' and his other Mustang without it.  This view is from the top down standing behind the passenger side of the trunk:  Look at the top fold line on the fender and then out to the fender flare.  The difference is clear.

I think that's subtle, but nice.  Here's a front view of the nearly complete car and it's honestly hard to tell.  This is with a 9.5" wide wheel with a 275 mm tire.

He did this with a quarter skin, not the complete panel, so I started wondering how to pull it off with a complete panel and not having a full length weld seam along the top of the fender.

Here's where he cut his out:

The Plan

My first step was to look directly next to my '65 over at my '67 fastback project car.  That wheel house is already there and the fender flare looks much more dramatic.

'65 flare:  This comes in at about 0.375".  I think my camera angle was off.

'67 flare:  This came in at  1.25"!  That's 7/8" more than the '65

So now, what if I could graft a '67-8 flare onto a '65 quarter panel with the '67-8 wheelhouse underneath?  Chances are the quarter panel wouldn't have to 'stretch' as far as the light blue car above since half of the 'stretch' is coming locally right at the flare.  Plus, the flare already looks like a Mustang flare so it may not look to far out of place.  Who knows, maybe some people wouldn't even notice.

Here are the two flares in a wider view:

65

67

Here's my step by step thought process and trust me, I'm open to feedback on what I'm thinking of doing here.

Step 1:  Cut off the quarter panel, but just the skin at this step.  Leave some reference points in tact, so if I overlay this piece on the complete panel, it's obvious how to line it up.  Perhaps the front door edge would be a good reference point.  SAVE THE ORIGINAL QUARTER PANEL!

Step 2:  Cut off the rusty, original '65 outer wheel house.  Toss in scrap bin, celebrate less rust on the car.

OK, so now the car should look like this:

Step 3:  Fit the '67 outer wheelhouse to the weld flange on the inner wheelhouse.  Temporarily clamp it in place.  It should look something like this:

Step 4A:  Using the saved, original quarter panel, put it back on the car and figure out how much needs to be cut back to let the outer wheelhouse clear the fender.

Excuse the MS Paint work, but here's the idea:

Step 4B:  Now working with the new '68 fender flare, fit it up to the wheelhouse and fender.  Determine a cut profile so that it will meet up reasonably well with the '65 quarter panel.  Cut both to size, so that they fit and would be ready to weld.

Step 5:  Laying the loose original quarter panel on the new full quarter panel, transfer the cut edge profile onto the new quarter.  Shift the line over about 1/2"  towards the wheel lip to leave extra material on the new quarter (measure twice cut once), so final trimming can occur later.

Step 6:  Cut off ALL of the original quarter panel in all locations so that the new complete panel can now fit up to the car completely.

Step 7:  Fit all three new pieces together on the car, trim, massage, etc.  When I'm sick of doing it, walk away, take a break - it can be better!  Come back a day or two later and keep going!  The goal would be to do a butt weld between the flare and the quarter.

Step 8:  Weld on the outer wheel house

Step 9:  Weld on the quarter and flare

Step 10:  Body work the flare to smooth out the transition between quarter and flare

Step 11:  Rinse and repeat on the passenger side

Measuring Shaun's car with the inner wheel house moved over already and adding 1.5", and I think there would be 14" of room for tires.  That's plenty to fit a 10" tire and have room left over for sidewall flex.

OK, so who thinks this is crazy?  Anyone think it's a good idea?

2012: 331 Stroker Ignition Choice

So now we'll skip to present day.  As I mentioned before, last summer saw the 302 in my coupe fail both head gaskets.  This has snowballed nicely into a completely new engine.  The only thing I'll be re-using is the carb (650 Holley double pumper), intake manifold (RPM Airgap), roller rockers, and the oil pan (Caton baffled road race pan).

My stock distributor in the 302 seemed to be cutting itself into the block below the cam gear for some reason.  I decided to look into what ignition systems were on the market.

The search begins

I had previously run a used MSD 6AL system with the Pro-Billet distributor.  I had strange misfire issues that I could never chase down.  Then when I removed those parts for a stock style system with a Pertroix drop-in points replacement - the problems went away.  I think the problems were either using incompatible plug wires, having the advance curves wrong, or even the fact that a MSD Pro-Billet doesn't have vacuum advance at all.  At least with this Crane distributor I can lock out the vacuum side or choose to use it.

Something always rubs me the wrong way when it is common to see a back-up MSD ignition box in race-cars.  These parts shouldn't fail.  With OEM reliability for these types of components skyrocketing since the MSD 6AL box came out, I was ready to see what else was on the market.  Some of that reliability has to be trickling down to the aftermarket.

Those old MSD parts are long sold off, so I had to look around anyway.  I saw that Classic Motorports Magazine (sister to my favorite Grassroots Motorports) used a Crane Street/Race Distributor in their 347 crate motor in their barn find '67 GT350.  What drew me in was the dial adjustable vacuum and mechanical advance curves.  No more springs and weights, no more vacuum diaphragms and hidden allen screws requiring me to count numbers of turns.  I wanted to dig in and learn more about how this distributor operates.  I like what I found.

Here's a look under the (positively screwed on!) distributor cap.  I haven't found similar detail online peering under the cap.  I will say that I'm pretty impressed with the build quality.  The bearing play is nearly impossible to detect by hand.  Everything fits very snug.  The cap does screw on so there is no play once installed.  Wiggle a stock distributor cap around on a 302 Ford once and you'll notice multiple degrees of play!  That can't be good for anyone.

The optical trigger can be seen near one tooth on the timing wheel

Here is the vacuum port for the built-in MAP sensor as well as the Mechanical and Vacuum Advance selection dials

Something I always wanted to do was to take a stock style distributor and test it on a distributor bench to understand the actual advance curves.  I always knew there were adjustments but I just didn't understand how effective the 'knobs' were.  I wanted data, but distributor benches are pretty rare in this day and age.  MSD publishes similar curves as Crane did here for their spring and weight kits so my methodology below could be applied to their distributor.  Going to the aftermarket with to get published data is the only answer.

Looking around online led me quickly to the Crane's website to find the PDF instruction sheet which has all of the advance curves detailed (pg 5 & 6).  They graph them one at a time and I wondered how they all overlapped, so I put all the data into Excel.  Luckily, once I got the box in my hand the actual data point values were printed on the box.

The first thing I noticed was that the total mechanical advance was different in a lot of the curves (20, 24, or 26 deg).  Typically Small Block Fords like about 36 degrees of total timing under wide open throttle at the high RPM ranges (assume 3000 rpm and up).  That means if we want to switch mechanical advance settings while keeping total timing constant, a shift to the base timing will be necessary.  Let's dig in and find out how much.

Since wide open throttle by definition means near zero vacuum, we can ignore any effect from the vacuum advance curve, that will only matter at lighter load and part throttle.

I'm going to assume we're working with Mech Adv setting #7

So:  Total timing (36 deg) = Base timing (x) + Mech Advance (26 deg)
So:  X = 36 - 26 = 10 degrees of base timing

Working that out for each of the settings we get the chart below:

Now we know how to set the base timing depending on which mechanical advance curve we've selected.  These numbers all assume 36 degrees of total timing, so if you want more or less total timing, these values would need to shift slightly.

By changing the base timing, it also changes how we should really be looking at the Mech Adv curves.  Instead of all of them starting at zero degrees advance and ending up with a different total advance, we should be lining up all of the curves at the same total advance.  In this case that's 26 degrees which is the max value in Curve 7.

Here's what that looks like.  There is less variation in the curves when you look at it this way.

Putting it all together

How do the vacuum and mechanical advance curves interact?  How would we know what to change during the tuning phase and which settings to change to?

First, from the guidelines in the PDF you have to pick which settings you want to start with.  It'll be different for everyone, so I won't go into detail.  I may start with M7 and V2, this would mean a Base setting of 10 degrees, so let's assume that for now.  It turns out that Classic Motorports used M7 on their 347 and my engine won't be much different.  Sounds like a great starting point.

Modern computer controlled cars are controlled by a lot of calibration maps.  Basically they are tables with two axis.  In this case we have engine speed (mech adv) and engine load (vac adv).  For each of these points with the curves that we have we should be able to figure out what the total advance is, this will be the value in the center of the table.  The equation for each point is:  Total = Base + Mech + Vac.  That's it.  Let's look at what that table looks like for what I'll call 'B10 M7 V2'.

Even though this is not computer controlled, it can help to think about it that way.  Afterall, it's still an engine.

I used Excel to calculate the advance in each cell of our table.  I used a the 'trendline' feature in the Excel graphs to spit out the equations for advance in the non-flat sections of the curves.

For B10 M7 V2:
Base = 10 degrees
Mech Adv deg = 0.0144 * (RPM) - 14.4                  From 1000 to 2800 rpm   For M7
Vacuum Adv deg = 3.2 * (Vacuum in-Hg) - 9.6        From 3 to 8 in-Hg   For V2

Notice how the base timing (10 degrees) shows up in the columns marked 'No Mech' and the rows marked 'No Vac Adv'.

The engine will idle around 750 rpm, but with lots of vacuum (12-18 in-Hg).  This is the bottom left of the table. In that case, notice that the advance is 26 degrees.  This is why you have to disconnect the vacuum line, or in the case of this Crane distributor zero out the vacuum advance dial, otherwise you'll see 26 degrees with the timing light, instead of confirming your base setting of 10 degrees.

Once the engine is under load and running down the straightaway, vacuum will drop (move up in the table), engine speed will increase (move to the right in the table).

Let's mark up the table and see what it looks like:

You can see that the middle box, the one where the grey sections of 'Mech Adv' and 'Vac Adv' meet is what I called 'Transition' on the chart.  This is the main driveability range, so this is where adjustments to the dials on the distributor will have a major impact.  Both of the equations above will impact the timing in this transition range.

Also, remember that unless something terrible is happening (detonation), you probably don't want to change the total timing (36 deg) in the WOT box.

How to understand the impact of changes

Now let's look at what happens when we start turning knobs.  To simplify things I'll just look at the center box (RPM:  1250 to 2750; Vac: 4-8 in-Hg).
Let's just turn the vacuum knob to #3  This takes out 4 degrees of vacuum advance, from 16 to 12 degrees, from vacuum alone.

Plotting both center zones:

I calculated the difference on the bottom right so that you can see that we really just changed the values horizontally one line at a time.  In other words it's the same difference going across the table, which should make sense if you look back to the entire table above at the grey zones where the equations take over.  We only changed the vacuum dial.

In this case, our base timing didn't change (B10), but since the vacuum advance did change from V2 to V3 that means the total amount of timing at idle will change.  I didn't plot it, but you'll have to take my word that it drops from 26 degrees to 22 degrees with V3 in that bottom left-hand dial.  This could drop engine speed slightly, so that might need to be adjusted.

Doing the same with the Mechanical Advance dial, you can see that the difference is now in columns instead of rows.

Here the base timing DOES need to be changed by the tuner from 10-16 in order to keep total timing at 36 degrees.  That means that idle timing would also go up 6 degrees from 26 to 32 degrees.  This will likely speed up the idle a little, so you'll have to adjust the carb idle speed down.

Using this method you could also look at what happens when you change both knobs.  

Let's go from B10 M7 V2 and change to B16 M8 V3

Note that your idle timing would only change from 26 to 28 degrees when keeping 26 degrees at WOT - AFTER you changed your base timing from 10 to 16 degrees.

Notice now that we've changed the mech adv and vac adv, the difference table has a diagonal color symmetry and that we've basically taken out  timing from this driveability range while adding 2 degrees to idle and keeping WOT the same.  Cool stuff.

How do we know which way to go?   Data, data, data!

In theory you could plot one of these tables for all of the different advance combinations this distributor has.  At this point we have TONS of data.  I wonder how much it would cost to run a distributor bench for this long and get the data for a stock Ford distributor?  I don't want to find out!

If we can take the next leap and understand which cell in the map we're having trouble with then we can look at all of our tables and see which settings might add or remove timing there while having the smallest impact on idle timing.

It works just as well today as it did 30 years ago to hook up a vacuum gauge and look at a tach while you're driving around.  Do a single gear pull to find a trouble speed/load point and look at your gauges to see what point you're at.  If you can get lucky and identify one steady state point you could repeat that driving event a couple of times with different curves selected each time.  Does more or less timing help?

The next step is to embrace a little more technology and record what's going on with the car.  This keeps your eyes on the road instead of gauges and takes the stress away from having to remember what happened.

I'll get into what I do in this data collection area in a later post, but here is a quick link to a new self-logging wideband that I just saw come out:  AEM

If you take a look at the PDF instruction sheet you'll see that it has it's own on-board vacuum sensor, so you plug a vacuum hose into it.  Then it has an RPM input as well as recording it's own wideband oxygen sensor.  It records the last 3 hours of run-time data, so you can go back and see how the speed and load were changing together and then trace your operating points onto the full speed/load timing maps above.

From my autocross days, I already have a stand-alone data logger recording:

• RPM
• Vacuum (GM 1-bar MAP)
• Throttle position (yes, on a Holley)
• Wideband
• Vehicle speed, acceleration, GPS position

I'll cover this in another post and also discuss how after a little time with the data logger, a throttle position sensor, and the vacuum data  it can really open your eyes to what's going on.  What's going on with the carburetor?  How far are the accelerator pumps are open?  Which circuits of the carb are functioning at any given time?  We can do all of this while comparing to real-time wideband data.