Porsche 914 GT Flare Project
Click on any image for larger view. You may have to hit reload on your browser for updates.

When we first attempted to reproduce the 914 GT flares, we imported several hundred factory flare sets from Germany at a reasonable price. After the initial purchase, the factory increased the cost monstorously so we could no longer afford to purchase. The sets were instantly sold and everyone seemed happy. three weeks later the problem began! Customers called reporting receiving two of one side, three of one side, or three fronts and one rear, etc., although they received four distinctively factory labeled parts! Since the flares themselves looked very similar, no one could tell which flare was which except by actually placing them against a car and comparing. There was no pattern to the mislabeling so each had to be placed, determined and relabeled. After this tedious process, and all of the exchanges, the mess was finally straightened. Which flare was which remains to this day with the factory sets, as they still are not labeling them properly! Rest assured our prefect repros will be labeled precisely so you will be guaranteed a matched set! Thus we have determined our project shall be known as, "The Flare Which Project."

Order your set of four today! Click here

After a long delay due to circumstances beyond our control, AutoAtlanta has now completed the 914-6 GT flare project. Every one loves the thought of the 914 equipped with the boxed 914-6 916 flares but few could afford the rare and expensive factory flared cars. Steel flares from the factory were expensive and now in their third reproduction, ill fitting and outrageous at $2650 a set, and fiberglass although cheap, is hard to install and does not bond to the metal well enough to last.

What better than a set of real steel flares reasonably priced, an exact reproduction of the factory originals and enable the 914 owner to achieve the look while directly bolting on a wheels as wide as eight and nine inches!!

We have done everything possible to make our flares better than the factory originals as we have spent much time smoothing them and with the lend of Brumos Porsche in Jacksonville of an old original set, our flares now fit better than the originals.
December 23, 2003

We managed to unearth the original measurement drawing from the GT that these flares were cop ied from!
May 21, 2003

Our first customer installation is being performed by Randy B, in MI. His 1973 914 2.0 is pictured here during installation.
May 13, 2003

Now that we have been able to finish both front and rear flares it is time to explain the process and success of how these were made. First of all, yes the factory still makes these parts - but whoever priced them must be nuts $2,600.00 per factory set! Who are they kidding. None of the factory sets were trimmed the same.

Even like parts were trimmed different in all the sets we tested. There were many wrinkles on each piece and evidence on lots of grinding and hammering all over the parts to make them look right. There are still many wrinkles all over the factory parts. What we did was took the best piece, did some repairs on each and used them to digitally scan. We extended the scanned parts on the computer screen by 3/8” to 1/2” depending upon the area around the part. That way when we stamped the parts we are able to give the customer more metal to work with.
We first scanned these at a company in Detroit Michigan called Capture 3D. They have a several hundred thousand dollar digital camera scanning system that can capture even the slightest change in shape on the parts, even a human hair if it happens to be there shows up on the scan. Then this is imported into our computer CAD, (computer aided design) system where we can work on turning this data into a mold that looks like the dies we wish to produce. We must remove the imperfections from the part and in the case of scanning the factory parts there were very many imperfections that needed to be removed and the area smoothed out. This took many days to do properly. We only scanned the best flare from one side on the car on both the front and rear. Once these were cleaned up and looked good we make a mirror image of it on the computer screen. We then turn them around facing each other as you can see on some of the pictures of the molds. We must space them far enough apart to leave room for trimming as well as enough metal for the bottom of the flare where it wraps under the body and attaches. Once that is done as shown previous, we cut an oversized Styrofoam model that goes to the foundry to be cast in steel. It is made oversized because metal shrinks as it cools by 1/8” per foot of material in all directions. It is critical that the model be sized properly or it will shrink too small to be able to be used. Once cast, it is put on the mill to be cut to its exact shape and size. These are so large that we cannot cut the whole die at one time even on our biggest mill. We have a mill that will cut a die 40”Deep, X 80”wide, X 32” tall and we still could not cut these all at the same time. They had to be cut the first 2/3 with three different sized cutters for about 40 hours and then turned around and the other 1/3 of it cut for about 25 hours. We have to put alignment holes in each part in order to properly turn it around and be able to set it up 180 degrees and be able to cut it.
Each time we do this it takes many hours to turn it around and line it back up. This was done for each half of each die for a total of four halves. Also each different tool we use to cut these requires a new tool path generated by the computer and each one can take many hours to produce. While we are cutting with say a 2” cutter we are producing tool paths for a 1” cutter because it can take 6 to 7 hours to produce. This is even with a 2.4 GZ Pentium 4 machine. These dies have over 200 Megs of data each. Sorry for those of you who could care less about gigahertz or Megabits. Anyway, after this is done each die half must be fit to the other mating half and tested or spotted together to make sure there are no tight spots. There always are so this is important. There must be clearance for the thickness of steel you are using. During the machining process we must machine extra metal from one of the dies to accommodate the steel thickness you plan to use. Aren’t computers great! We can tell our cad system how mush extra to remove. By the way, CAD means, computer, aided, design. We also have a CAM software system, computer aided, machining, as well.

The spotting of the dies is done in press and must be done carefully. If the dies are too tight together and you mash them together under 250 tons or 500,000 pounds of pressure, well, lets just say they do not want to come apart very well. Can you imagine spending over 100 hours and thousands of dollars in materials on a couple dies then have one of your employees jam it together under many tons of pressure. Well, this did not happen, as we are very careful to be fully involved with this process. It can and does happen however and the dies are very difficult to get apart. This process must proceed slowly and carefully and as I mentioned takes at least a couple days of sanding and polishing. We test fit small pieces of the steel we plan to use during this process and work our way up to larger pieces. We cannot test a full sheet of steel during this process and these dies need a binder ring around them to prevent the metal folding over itself.

A binder ring is a steel ring with the center cut out and fits over the bottom die and sets against the top die. Steel pins stick up out of the press on an air cushion and the binder ring sets upon these pins. The sheet of steel you use is placed upon the binder ring, which is setting above the bottom die on the air pins. The top die is bigger that the lower die, and has a shelf or ledge around it that matches the binder ring. When the press closes the top die pinches the steel sheet between the binder ring and the upper die. This binds or holds the steel all the way around and it stretches the steel over the bottom die as the press closes. By doing so it hold the steel and helps to prevent any wrinkles or overlapping of the steel as it flows over and around corners to make the shape of the flares. Because of the shape of these flares this was an absolute must to have. The trick is to balance the air pressure in the air pins to give you just the right amount of holding of the steel. Too little and the steel will flow into the die too fast and it will wrinkle badly. Too much holding and you will tear the steel in several places especially where it draws over corners. We threw away 15 to 20 sets of parts during the tryout stage of perfecting the holding pressures.

Also with these parts it is impossible to make these with one draw all the way down. These parts must be stamped about 85% of the way, then some of the center of the part between the wheel openings must be cut away, then the press closed all the way down and then the parts is stamped again. If you go even .040 or .050 thousands of an inch too far on your first draw the part stretches too far and tears or gets a stress fracture.

Our goal throughout this entire process was to make these flares better and less expensive than the factory parts, which sell for $2,600 a set. We have indeed succeeded in both ventures. We did have problems, some of the same ones the factory encountered but we were able to spend some time and try different things to overcome most of them. We had to compromise once in the area of stress wrinkles and stretch marks that the factory also had problems with. During the stamping process we had to partially cut the parts in the center. This relieved the part and allowed the metal to flow properly. We also had to decrease the binder pressure from 105 lbs. down to 50 lbs. for less holding pressure and allow the metal to flow easier. This was done to stop a problem on the face of the part directly over the wheel arch area. We were drawing the metal on the second hit and it was dragging the already flanged wheel arch edge up into the face of the flare so a line could be seen easily all the way around the wheel arch. There was the old flange line about 3/8” up onto the side of the part. It looked bad, but the rest of the flare looked great where it met the body. Well, this was a problem the factory had because all of the factory parts we had which were about 3 sets all had this problem. They decided to grind it away. Not acceptable. This is when I lowered the holding pressure on the binder ring to 50 lbs. and trimmed out less of the center of the part. After destroying many parts we were able to draw a flare that was nearly perfect around the wheel opening with no extra flange line around it but we were getting some small wrinkles or stress lines at the top of the part where it attached to the body. The rest of the part was great. Let me ask you, where would you rather have an area to work on a little, at the top where you have to weld the part to the body anyway or right on the side of the part in plain sight, where you did not intend to spent time filling or sanding out imperfections? Well we decided it for you. There will be some stress marks or wrinkles at the top of the flares, but the face of them you will be very impressed with, especially those of you that have seen or installed the factory parts.

The last stage is to use thin rubber pads to stamp most of the wrinkles at the top of the parts. We hit these flares about 6 times from start to finish to make sure you will be pleased with the fit and quality.

These flares are by far of better quality than the originals. Who knows, maybe we should be selling flares back to Porsche from now on. Automobile Atlanta’s flares are better quality and made a little heavier and are of better quality steel. The steel used is what the auto companies brag about using in their new cars. Zinc coated on both sides so they resist corrosion and they accept paint better than bare steel.

It would be cost prohibitive to try to make wrinkle free flares as it would be about $200,000.00 or more for tooling in able to accomplish that. Sorry for the long drawn out explanation but we thought it would be interesting for you to take a small trip down tooling and stamping lane to learn a little more about the flares you are interested in purchasing and what it takes to give you a quality product.

April 21, 2003
These are shots of all four flares and then just the rears. These are the first good set we ran this past weekend. There is more work to do on the dies and blank size which we will be doing tonight. If all goes well we will be able to start shipping full sets this Wed. or Thurs. It takes a while to stamp and trim these as they need to be hit in the press about 4 or 5 times for different processes. We first have to stamp these about 80% of the way closed. Then we must trim out the center between the two flares. Then the press is closed farther and the parts are stamped again. Then we final trim the parts and hit them again because the trimming can alter the shape some. Then we adjust the press up and put rubber pads where there may be some small wrinkles or stretch marks in the part and we hit it once more. Then if there are any marks that do not want to come out, we tap them out with a body hammer and dolly. So you see there is a lot of time and work that goes into the process of these flares. The front ones as I mentioned are ready to ship and we are now working at getting the rear ones to that stage.
March 27, 2003

This is a picture of the front fender flare that was stamped last night, finally. The press that was rented broke down twice and had to be repaired. We are getting some small wrinkles exactly where the factory did at the very top but it is better by far than theirs everywhere else. We are working on the wrinkle problem to get rid of them entirely if possible. It is already better than the factory ones after only having stamped 5 sets so far during our tryout stage. More to come.
February 18, 2003
The pictures below are from Sunday at nearly midnight. The front flare Post half of the die is pictured. We are still doing the finishing machining. Our cad machine is a 2.53 Ghz machine with 1 GIG of ram and it still takes 4 to 7 hours to produce tool cutting paths before you can use them on the mill. Then the mill takes them and spends another 10 to 12 hours cutting the surface with the data. We are in the final stages of this with this die and it should be ready to put into a press by Thursday this week.
February 15, 2003
We have the second half of the front flare die in the mill. Time is 4:30 pm. The cad machine needs 5 to 6 hours to create the tool paths we need to cut this half of the die. It will work the rest of the day and tomorrow we will come back and run it all Sunday afternoon and Sunday night. It will take about 20 hours to do the first cutting of this tool so it will still be running when we come on Monday morning. It will then take about 12 hours to do the 1/2" passes then this one should be done. It will not need the 1/4" passes because of its shape. The other half did. If all goes well we will be spotting the dies together by Tuesday afternoon or Weds. I hope to be in the press with this die sometime late next week. For this set of dies I ordered a binder ring to hold the steel in place while stamping. It is needed for this part and will be needed for the rear's as well. I could make a ring than fit both front parts and one for both rear parts by just turning it over. No choice, since these will not form properly with out one for the front and one for the rear. They are different sizes so two are needed.
February 14, 2003
Febreuary 10, 2003
Back from the foundry pattern shop where they just finished cutting the foam models for the rear flares. The models will go to the foundry to be cast next. They will need it for about 10 days or so. The front flares are now cast but they need a few days to cool off and be cleaned. Soon to start machining. These dies have been designed with every possible advantage to assure that the dies will work properly and make good parts. We are doing everything possible to make them better than originals.
January 27, 2003
We first took the optical scanned data which was millions of points and converted it to millions of small triangles in a cad system. From this data also we were able to draw lines across the data to make a grid of lines going two directions. The parts have imperfections in them from years of setting around and draw marks from the stamping process. These can be corrected in the cad system by smoothing out the lines where you know it is not right. After many hours of work on one scanned part we then make a mirror image in the computer and move the two close together. The are then rotated and filled between to make the best shape possible for metal to draw over top of them. This is done because drawing two parts at a time will draw better than just doing one. We extend the surfaces about 3/8 to 1/2 inch all the way around to allow for trimming the part after it is stamped. You will notice a crossed area in the middle of one of the dies. This is for strength as this particular die is smaller than the others and needed this area to keep it together during the stamping process. You will notice the centers removed from the other dies as well. You will not see it but we have hollowed out some of the bottom of each die leaving at least 2 inches of material at the surface in all areas. This also lightens the die but also makes for less shrinkage because it has less mass. Speaking of shrinkage, we machine Styrofoam molds to take to the foundry and they have to be made oversized to allow for shrinkage of the steel as it cools. Metal shrinks an average of 1/8" per foot in all directions as it cools. The more mass there is the better chance of warpage and low spots in the dies. This is why we cut hollow pockets called cores into the bottom or sides of the molds. They will still be strong as we leave a web between the cores for strength. The molds are made 1% oversized plus an extra 1/4 to 3/8" extra for machining purposes.
The Styrofoam is put into a sand cast box and sealed all the way around with sand leaving a couple fill and vent holes. The foundry will then pour liquid ductile iron at about 2,000 degrees into the sand box. The liquid steel totally vaporizes the styrofoam and conforms to the shape that the foam was in. It is left to cool which takes many hours and on top is left a sprue area with excess steel poured into it to allow metal to flow into the die as it cools. This along will coring helps prevent low or hollow spots in the casting. This entire process must be done with each half of each die for in this case a total of four times. Four molds must be completed in the cad system and inspected to make sure nothing has been forgotten and four machined out of foam and four cast into boxes and made into steel dies. After many many hours of cad and model work you will get back 4 dies that look much like the pictures we have sent you. Now remember each one had to be made oversized so each must go into the mill and have about 1/4 to 3/8" machined off it to be of proper size and shape. When the dies come back from the foundry they will be warped and a little uneven so they go into the mill and the bottoms are fly cut flat so that we can turn them over and have a flat surface to work from. This also must be done 4 times. Each die will then be cut with a 1" ball end mill, then a 1/2" and then where needed a 1/4 to 1/8" to insure proper radius's. Remember this is done to each die. Some of you by now are asking why go through so much hassle and work. Would it not be less is we just carved it out of solid steel blocks. The answer is no. The steel blocks are much more expensive to buy and the hours of machining each of them much more. Each block for this project would weigh 3700lbs, and 5,000lbs for the front flare dies and more for the rear as they are bigger, 4600lbs and 6300lbs. That totals 19,600lbs of steel blocks much of which would have to be machined away which takes many many hours. That much steel alone would cost about $12,000 by itself not including the machining and handling problems moving heavy iron around causes. The castings cost more per pound to cast but will weight considerable less, about half as much weight and we will only have to run finish cuts to it because it will only be a little big but have all the right shape. One mold is going to the foundry for casting. Once this is done, we will be cutting on both halves for about a week and then be able to spot them together in one of our presses. That takes about two days to properly fit them together and then we can try them out in a larger press with a binder ring and proper blank holding. We should be able to produce parts within a day or two of tryouts and then start production of finished goods. My guess at this point is I will be able to have samples of the front flares by the end of February. The rears will follow soon after but will take a couple weeks more. While we are finishing and trying out the fronts we should be getting back the rears from the foundry and start machining them. The same process starts over with them until they are able to be stamped properly with a binder ring for blank holding.
December 18, 2002
Shown are pictures while at Capture 3D optically scanning the flares (these are the same people who scanned the new GT-40 project for Ford) and doing some preliminary hammer and dolly work. The scans were taken by an Atos scanner which is able to capture 1.3 million points in seven seconds, so the original flare can be captured in every minute detail (including scratches!!) without distortion. We also at the same time are scanning other 914 parts for near future projects (notice the exterior rocker panel! These will be a real deal!)

With these detailed scans we will be able to produce a CNC (computer numerical control) program directly from the scan data and use a mill to recreate our model!
Stay tuned for more, and get on the list for your brand new set of 914-6 GT factory look flares!
| Back to Porsche Life | Order your set of four today! Click here | Message Board |