By on July 10, 2012

One of two circular looms on the planet. 12 layers of seamless carbon fiber are woven into what will be part T3-3RH, part A-pillar, part roof support

Yesterday, we heard how the LFA really was born (in a bar, where many good ideas are born and pitched,) and why it is made from carbon fiber. Now, we are in front of the cleanroom, and while our little group is suiting up, let’s use the time for a quick course on CFRP. 

The basic principle of Carbon Fiber Reinforced Polymer, CFRP for short, is not new. It dates farther back than metal. CFRP is a composite, made from two completely different materials that are joined together to give a much stronger material. Straw and clay was such an early composite. Concrete is a more recent one. In the case of CFRP, carbon fibers are combined with epoxy, the polymer. Sometimes, the material is also called “carbon fiber reinforced plastic,” but the end product is far removed from what usually comes to mind when we think of plastic.

Chief engineer Tanahashi shows a pre-pregged carbon fiber mat

CFRP has an unsurpassed strength-to-stiffness-to-weight ratio. CFRP also comes at unsurpassed cost. It entered car making with race cars, where money is no object. As we enter the cleanroom through an air lock, it becomes evident why the stuff is so expensive.

Basically, there are three different ways of making CFRP: Pre-preg, Resin Transfer Molding,  (RTM), and  Sheet Molding Compound (SMC), listed in the order of strength and expense. The LFA uses all three methods, depending on the required characteristics of the CFRP parts.

This reporter is being vacuumed to protect the LFA’s carbon fiber from filth and grime

The strongest parts of the LFA are made in a clean room. Our hair is covered. We wear long white coveralls. Booties go over our shoes. Someone vacuums me from top to bottom.

Cleansed, we enter the cleanroom

We enter the cleanroom through an airlock, and we are in pre-preg central.

Pre-preg is the method to make the strongest type of carbon fiber available. “It is also the most expensive,” says Masahito Miyoshi, who leads me through a maze of machinery. Miyoshi -san is in charge of a by-product of the LFA, a mountain of paperwork. He also has the honorary title of the LFA’s ambassador.

Working in the cleanroom is choice duty: The room is not just dust free. The temperature is kept at 22 degrees centigrade (72 F), the humidity at a constant 60 percent. A cool paradise compared to the hot and humid Japanese summer.

Pre-preg is not what some may think it is. Carbon fiber mats that were pre-impregnated with resin begin a metamorphosis into the LFA’s most critical parts.

A plotter/cutter divides the pre-pregged carbon fiber into hundreds of parts

The carbon fiber mats are bought from a supplier and wait in a giant freezer until they are used. The mats have two layers of carbon fabric, the threads of the fabric layers are oriented at distinct angles.

The pre-pregged mats come to rest on a large cutting table the size of a small Japanese apartment. Vacuum sucks the mats into place. A computer-controlled plotter first draws parts numbers on the sheets. Then, the plotter withdraws its pen, brandishes a rotary cutter and in 20 minutes, the mat is dissected into a giant puzzle that will soon do duty in LFA number 424.

The dash starts with this mold

A few steps down the air-conditioned cleanroom, workers produce a dashboard. Under the hawk-like eyes of a foreman and by gloved hand, the workers layer 335 pieces of pre-pregged puzzle into a mold the size of a, well, dashboard. Sometimes, they use an industrial-strength hairdryer to form the puzzle-piece into its desired shape, and to avoid any air bubbles between layers.

335 pieces are assembled into a dashboard, a full day’s work

Like a colossal wedding cake, 13 layers of pre-preg are methodically assembled into what will be the dash panel. The layers are mated with urethane core. The cores have threaded aluminum inserts, later, they will be the counterparts for bolts.

Each piece of backing foil is accounted for

Each of the 335 pieces of the dash is checked off from an evidence sheet. The sheet gets signed by the foreman, and then signed again by a quality controller with wasp-like yellow stripes on his hat. The backing foil of the puzzle-pieces gets pulled off and is discarded, but not without being recorded on the evidence sheet.

“If they forget the backing foil, nobody would see and notice,” explains Miyoshi. “Until perhaps in an accident.” Every piece of backing foil is immortalized on the evidence sheet before it is thrown away.

One works, four watch

It takes eight hours to hand-layer this dashboard. Once done, the finished wedding cake dash is covered with a special vacuum foil and baked for another eight hours.

With pressure and heat, parts cure in the autoclave for eight hours

The baking happens in the autoclave next door. Picture it as a garage-sized pressure cooker. Pieces in the autoclave are covered with foil, then vacuum is applied that presses the layered piece into its mold. Two bars of pressure is applied to the autoclave, and at a constant temperature of 150 C, the parts are cooked to perfection.

The autoclave works two shifts per day

That autoclave is in high demand, and it is the only station in the genesis of the LFA where two shifts are working. During the day, a dashboard and other parts cure in the pressure vessel. At night, the side members of the LFA are being baked. This limits the production speed to one LFA per day, and the profitability of the venture is limited to forget about it. In the same time one CFRP dash is completed, a conventional machine spits out more than 1,000 plastic dash panels, far less rigid, far more weighty, but far less costly than the LFA part. Perfection has its price.

Crazy like a loom.

This time, it’s for our protection

Barely out of our white garments, we don masks again. This time, it is for our protection. Breathing microscopic carbon fibers could be hazardous to my health, I am told, as we approach a wondrous machine. It reminds me of the circular loom that was invented at Toyota in 1906, and became part of the foundation of Toyota’s later fortunes.  The new circular loom weaves carbon fiber and could become an important part of Toyota’s future.

Tanahashi explains how the loom works

The new loom is called a three-dimensional braider. Only two exist on this planet, I am told, the location of the other one remains undisclosed.

Myoshi and Tamura inspect the (black) carbon sock over a (white) wax core

What looks like a long black sock goes back and forth six times while carbon fiber is spooled off 144 bobbins and braided over a core of wax that rests in a vacuum-sealed ABS pouch. Layer upon layer of carbon fiber is woven until the core is covered with 12 coats of fiber.

Tanahashi holds one of the world’s thinnest and strongest A-pillars

The finished sock-over-wax goes into a press between two molds, resin is added, heat and pressure are applied. Eight hours later, the wax is molten, and Chief Engineer Tanahashi can inspect part T3-3RH that will be part A-pillar, part roof support. The 12-layer seamless sock  becomes an embodiment of supreme strength and luxurious lightness. It is one of the thinnest and yet strongest A-pillars in the business.


The 3D weaver makes a seamless mat of carbon fiber …

Bumper and grind.

Next door is another marvelous machine. Normally, weaving is a two-dimensional affair. This machine weaves into the third dimension. It interlaces 32 layers of carbon fabric, with the layers on top of each other at varying angles, into a thick mat. Carbon fibers have been reinforced before by sewing them like a quilt. However, this can also weaken the quilt along its threads where the fabric is pierced by the sewing needle. The machine avoids this by weaving into the third dimension.

… which ends up as this crashbox

With the addition of resin, the carbon quilt is transformed into a part that is held here by the LFA’s Deputy Chief Engineer, Chiharu Tamura.

“It is used for the front bumper stay,” says Tamura, who hopes that the part will never have to prove its true characteristics. “It is a crash box. The vertical fiber gives it its superb energy absorption characteristics during a crash.”

A crash box is amongst the most important automotive parts for crash energy absorption. Situated at the front of the car’s frame, a crash box waits to be hit during a crash. After impact, it is the job of the crash box to absorb and to dissipate crash energy before other body parts of machine or man are impacted.


High speed camera picture of the crashbox in action. The inset shows its location

A conventional crash box crumples as if soda cans are squeezed by a giant hand. The LFA’s 3D woven crash box does not crumple. It crumbles. A black and white picture, shot by a high speed camera during a crash test, shows how the material behaves during a crash. The crash box fragments into millions of small particles, each particle requiring energy for its separation. This box can effectively bleed-off the energy of a crash until the force falls harmlessly to the ground as millions of black crumbs.

These days, a lot of the crash testing is performed in computers, and then, to be sure, it is performed again in reality. Modeling the behavior of metals is fairly straightforward. With CFRP, the engineer is overwhelmed by a multitude of constants that are the product of fiber quality, fiber orientation, fiber density, the resin used, the number of layers, the angles of layers, the manufacturing method, and many more. In their research, the LFA team first made small samples of materials, tested them, and then used their properties for the larger test. Nevertheless, “much more crash tests were performed with the LFA than with a regular car,” Tamura said. How many, he would not disclose.

Different types of CFRP are used for different loads

 Not all carbon fiber is created equal.

The pillar/roof support that comes out of the circular loom, the crash box, and many other CFRP parts of the LFA are made by another process called Resin Transfer Molding or RTM. With RTM, dry fiber is laid into a mold, liquid resin is fed into the dry fiber, 130 centigrade of heat and 3 bar of pressure are applied, prompting the part to cure. The cured parts are then machined to perfection using ultrahigh-pressure abrasive water-jet cutters, thankfully, away from our eyes.

RTM is one step closer towards mass production, but it still is a long way removed. RTM saves the cumbersome hand lay-up of pre-preg. There is no massive autoclave. However, the part still has to remain in the mold for eight hours until the resin is hardened. Eight hours saved, but still eight hours to go.

There is a third process called Sheet Molding Compound or SMC, but Tanahashi and Tamura give it short shrift as we walk through the LFA Works. With SMC, chopped pieces of carbon fiber, each about one inch short, are mixed with resin to create a high-tech version of paper mache. This mix is applied to a sheet, it goes into a mold, where it is cured under heat and pressure. This is the lowest grade of CFRP used in the production of the LFA. It is used for parts that do not require a large amount of strength, such as side panels, fenders, or parts of the rear area of the LFA. These are the only CFRP parts that are entrusted to outside suppliers.


A robot applies a rectangular bead of glue

A bond for life.

We follow Chief Engineer Tanahashi down a long corridor until we are faced by an unexpected apparatus: A robot.

Usually, most of the heavy work in car building is performed by robots. With the LFA, the robot has the job as a glue dispenser. Fed by a long hose, the robot applies carefully measured uniform amounts of epoxy resin adhesive to the many CFRP parts that make up the passenger cell of the LFA. The adhesive is not applied like the rounded bead we all know, but rather as a flat strip with square edges.

In a jig, the parts are fitted together for a one piece passenger cell. This jig is the only part that is kept from your eyes, I am not allowed to photograph it. Bonded together, the different parts of the passenger cell effectively become one, extremely light, but yet superbly stiff and strong.

Building the body of one LFA takes four days, we race through it in three hours. On the fourth day, there is what is called a body-in-white in the car making industry. Except that in the case of the LFA, it is a body-in-black. In a shiny, glossy black that we usually associate with carbon fiber.

When it is done, the people in the CFRP shop stick a picture of an LFA with a number to a whiteboard. 424 done, 76 to go.

Stay tuned for tomorrow’s installment of The Making Of The Lexus LFA, where we will follow the body to painting and assembly. 

Monday, July 9: From A Bar To Bar None. How the LFA was born, and why it is made from carbon fiber.
Tuesday, July 10: In The Clean Room. Where the LFA is made from the strongest and most expensive type of carbon fiber available.
Wednesday, July 11: Call Me Names. How the LFA really received its name.
Thursday, July 12: Balance Of Power. We watch the V10 engine go into the LFA.
Friday, July 13: Exam Week. We examine Chief Engineer Tanahashi about how the LFA influences future cars, and what will come after the LFA.

Get the latest TTAC e-Newsletter!

46 Comments on “The Making Of The Lexus LFA Supercar. An Inside Report, Chapter 2: In The Clean Room....”

  • avatar

    When I first glanced at one of the photos above, I thought they were hockey sticks (sorry, but I’m Canadian). Is this the same composite material that’s used in hockey sticks?

  • avatar
    Athos Nobile

    “The adhesive is not applied like the rounded bead we all know”

    For gluing the windscreens to the cars (or other glass) I remember the Sika or 3M representatives always recommended to avoid the rounded bead. Instead, they recommended the use of a triangular one. There’s even a special nozzle for that. At the end of the day, that one ensured a correct seal.

    Having seen how they made fiberglass cabins before, I could possibly picture the jig myself.

    The CAD rendering you posted even show where the adhesive goes, and it doesn’t take much to see the panel on which the robot applies the glue is the LH where the door goes.

    Nice write up. I am interested in this “How the LFA really got its name” and more importantly, “We examine Chief Engineer Tanahashi about how the LFA influences future cars, and what will come after the LFA.”

  • avatar
    The Doctor

    Thanks for sharing this Bertel – I’d always wondered how the “three-dimensional braider” built up the shape of the object it was weaving.

  • avatar

    Hopefully, the “crash box” is one of the first parts that will be installed in all Toyota vehicles and offered as an OEM part.

    I had to chuckle reading this line…”the profitability of the venture is limited to forget about it.” Typical long term “Toyota” thinking.

  • avatar

    Really fantastic content Bertel. Thank you for bringing it to us.

  • avatar

    The chart with the material properties is messed up.

    The x-axis must be *specific* modulus of rigidity (aka specific shear modulus), because steel’s shear modulus is ~3x aluminum’s (75 GPa compared to 26 GPa), but it is also ~3x as dense, so that’s how they can be so close on a chart. But that means the units are all off, because specific shear modulus is measured in GPa/(kg/m^3). And why use shear modulus instead of tensile modulus?

    The units are also off for specific strength. It should be MPa/(kg/m^3).

    There is so much variability in composites’ properties (volume fractions, directionality, matrix material, etc) that it’s nearly impossible to check those values.

    Also, CF composites’ properties are dwarfed by nanotube composites. However, I believe their costs are still in the Apollo program range.

    • 0 avatar

      I eagerly await the Lexus Space Elevator!

    • 0 avatar

      Now you want me to edit their charts?

      • 0 avatar


        I must say that with this series on the LFA, you have just hit automotive journalism out of the park. This quality of work, insight, and reporting exists nowhere else. Congratulations.

        I don’t know how everyone on your staff makes money (advertisers?), but I would not object if you charged a subscription fee for access by your loyal readers. Yes, it’s that good.


    • 0 avatar

      That chart is garbage, a marketing chart, not something you would show to anyone who knows materials.

      Lockheed is using a thermoset epoxy strengthened with carbon nanotubes in the F-35 wingtips, which are several times stronger than CFRP and also lighter by ~20-30%. Supposedly they developed a process which makes the CNRP cost much less than traditional CFRP, but I’m skeptical of that.

      All the stuff Bertel just detailed is twenty or thirty year old technology, with the exception of the 3-D braiders which have been around since 2005. The real revolutionary technology will be how automakers can manufacturer the composites for $25k cars. Lexus failed pretty hard on the LFA at that, selling $350k cars at a loss.

      • 0 avatar

        What’s the name of that Lockheed car that’s cheaper than the LFA?

      • 0 avatar

        Ah yes, the F-35, more expensive then the F-22 with f-16 performance, a really nice benchmark for affordable technology. It’s not that Toyota are revolutionizing material technology full stop, it’s that they are working – quite successfully – at developing know-how in composites for car applications, and that might prove to be a revolution. I’m sure that toyota could build an even lighter car if the could price it in the $150 million range, if they could deliver it whenever they felt like and if they could force the Japanese government into buying it and strong-arming other countries into buying it.
        I do agree that the facts and figures in this reportage smells of marketing rather the engineering albeit not as bad as the BMW M pictorial history abortion.

      • 0 avatar
        Robert Gordon

        “I do agree that the facts and figures in this reportage smells of marketing rather the engineering albeit not as bad as the BMW M pictorial history abortion.”

        What ever happened to the third part of that ‘M’ series article? Some of my colleagues were hanging out for that one….

    • 0 avatar
      Greg Locock

      Sorry redav, if I had read your excellent post before writing my one at the bottom, then I wouldn’t have bothered with mine. Obviously.

  • avatar

    Very interesting stuff. Thanks Bertel! My professional background is in relatively low volume one-piece woven curtain air bags (textiles) and high-volume rubber molding (not exactly CFRP, but similar in principle), so this article was particularly interesting for me. I’m looking forward to the rest of the week’s articles.

    Side note: it’s kind of amazing how many Lexus logos they have on their uniforms.

  • avatar

    More wondorous insights. Every time I see that loom, I swoon. As for the 3D weaver…I did not know men could build such things…but I shouldn’t be surprised. My already high regard for Japanese ingenuity proceeds apace.

    I also laud the LFA’s daddy (daddies?), for having the pre-preg CFRP will and epoxy-resin sticktoitiveness to keep this project alive through thick and thin. Like climbing a mountain because it is there, only it isn’t and you have to build it first!

    I only have one question (for now): how on earth did you get access to this magical chocolate factory?

  • avatar

    So RTM requires no massive autoclave(does this save on electricity bills?), and there is another circular loom at an “undisclosed” location. Come on Bertel, tell us everything about how Lexus is working on making a critical piece like an A-pillar, so that even more weight can be taken out of production cars.

  • avatar
    The Doctor

  • avatar

    These remind me of the three dimensional weavers used in ‘growing’ organs. I wonder how long it will be before someone tries to ‘grow’ a car? (And I’m not being facetious here).

  • avatar

    Wow – that is a giant middle finger to quarterly earnings.

    • 0 avatar
      juicy sushi

      I have to wonder how important the support of Akio Toyoda was to this project. When it’s your name on the building and you’re the largest shareholder, I have the feeling you get to play in the technology sandbox a bit more than others might allow.

      Bertel, thank you very much for these articles, and please thank the people at Motomachi for being gracious enough to share their neat toys with us. If I could find a wallpaper quality photo of the braider for my computer I’d be pretty happy. A screensaver of it in action would probably be even cooler. I’ll have to see if I can cook one up.

  • avatar

    I’d be interested to see how different the LFA production is from Avantador. Since that car also has some rather unconventional carbon shell/unibody vs the more traditional tub. And they also use some more “mass-production-friendly” technique to build a carbon car.

  • avatar
    Felix Hoenikker

    I wonder what the radar profile of this buggy looks like. Could it be the first stealth automobile?

    • 0 avatar

      No that honor probably belongs to this car I remember it from an old Benetton magazine that had some “interesting” things one could buy.

  • avatar

    This series is making the $375K seem like a bargain for the LFA

  • avatar

    Given the dangers of breathing loose carbon fibers, will the LFA have to be handled as hazardous waste after a crash?

    • 0 avatar

      Probably not since the fibers are still bonded to the epoxy at that point. However, you would need to be careful about sticking yourself with the exposed fibers at the break points. However, I wouldn’t stick my face near it and breath deeply, either.

  • avatar

    This is great.
    We shall have the “nucleo lab” seen in The Fifth Element in no time.

  • avatar

    I think those crash structures are where it’s at in moving carbon fiber to the mass market. On a F-1 car, they’re incredibly efficient at absorbing impact forces, because they deform along the fiber lines… if you ram a F-1 car into the wall, the nose doesn’t crumple, it disintegrate. High speed shots of this are incredible, you can see the impact forces tearing along the individual fiber lines.

    I picture something like the loom in the room weaving generic crash boxes, which are simpler than whole chassis tubs, which con on to a convey-er belt autoclave system. You can then build a smaller and lighter car around CFRP impact boxes.

  • avatar
    schmitt trigger

    First stealth auto?
    It might…
    But would it have to have those puny windows and the panels at bizarre angles much like the F117?

  • avatar

    Wow. Just wow. When I first saw pictures of the LFA I thought “Just another high end sports car.” This little series gives me a much greater appreciation for the vehicle. It also reminds me to never doubt Toyota. They may churn out coma inducing cack like the Corolla, but the company is innovating in so many other ways.

  • avatar

    This technology will never be useful in mass production. This is a ‘prestige project’ by a company that has more cash than sense.

    • 0 avatar
      el scotto

      Why not? Aluminum used to cost more than gold. Granted aluminum takes an ungodly amount of electricity to smelt. Perhaps someone will make a similar breakthrough with carbon fiber?

      • 0 avatar


        I’m not sure it’s as simple as that. Unlike just a “concept” car, the LFA became a “proof of concept” car for which immediate financial ROI was set aside.

        All really new technologies start out that way and may look silly from an accounting point of view, but efforts like this help Toyota come up a learning curve that will no doubt pay future rewards. Remember, the Japanese take a long view of things like this, something I wish GM had done before they just killed the rear-engined Corvair in the 1960’s; and killed the all-plastic Saturn recently. Each could have been corrected technologically for long term success, but met their demise at the hands of bean counters (and others) because of a need for better publicity or short-term profits.

        Battery electric vehicles (BEV’s) may be like that too. Currently they are languishing in the market, and despite my love of ICE’s, can you imaging what would happen if a Toyota-type company invents a polymer battery that is actually a coating a few millimeters thick that can simply be applied on the inside of all body panels; costs no more than aluminum; lasts virtually indefinitely; has an energy capacity to propel a car for 500 miles; operates at -20 deg F; and recharges in 1 minute! Yeah, it may look like an accountant’s worst nightmare initially, but the future benefits would be HUGE. And BEV’s would finally really have their day in the sun. To my dismay.


  • avatar

    I didn’t realize they could do lost-wax molding of carbon fiber. It seems like the part would crush under the pressure as soon as the wax melted.

    This is probably the secret way Santa Cruz makes it’s absolutely gorgeous, seamless one-piece carbon bike frames.

  • avatar

    Really interesting series, Bertel.

    CFRP has been around for decades, my 1986 AWD 1/8 scale RC gas car had an optional CF chassis. Problem was, when you hit something hard it shattered like that LFA crash box, while the standard aluminum chassis could be bent and tweaked straight again. Er, usually, that is.

    Just visited the McLaren MP4-12c website, and of course, they use CFRP as well, since they made the F-1 of the same stuff back in 1992. It’s a pity that the stuff is so difficult to mass produce, and that the only advances Toyota seems to have made is the 3D braiding circular loom, since I believe the 3D weaver is a commercial unit.

    The unfortunate death of a Lexus test driver on a road outside the Nurburgring while driving an LFA now surprises me, however. He had a head on with two factory BMW drivers. They lived, but he didn’t. So is that crash piece substantial enough, I wonder?

    In any case, It’s good to see that Toyota is keeping up with a “side” technology and their linkup with BMW to cooperate on CFRP matters is significant. BMW makes carbon fibers from scratch, as a You Tube video demonstrates; there is also You Tube video on that Toyota 3D loom.

    Fascinating stuff.

  • avatar

    Why couldn’t you photograph the jig? I guess I can guess why.

    What is the weight of the A-pillar?

  • avatar

    A remarkable amount of refinement for a car that gets KILLED by the Nissan GTR

    • 0 avatar

      Yes, comparative Nurburgring lap times would say that. But again, this whole project may be a proof-of-concept car for new technologies, more than just a fast car for sale. After all, a simple Porsche 911 GT-3 RS can beat BOTH the LFA and GT-R around the ‘Ring (when properly tested), and it’s a “lower-technology” vehicle than either. Please see my comment to “icemilkcoffee” above.


  • avatar
    Greg Locock

    Not too sure about that graph, it has at least one mistake. Aluminium alloy has about 1/3 the stiffness (rigidity) of steel.

    Ah, OK it seems to be normalised by density. The x axis is wrongly labelled. It should be “specific rigidity”.

  • avatar

    Is that prepreg layup part really a dashboard, or Is it a front firewall?

Read all comments

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • slavuta: Why do I need to listen to some Ellie Murray? for “week ending Dec. 10,”??? I went to official...
  • Lou_BC: “Last week, unvaccinated individuals were 31x more likely to be infected with COVID than boosted individuals”...
  • Lou_BC: “Last week, unvaccinated individuals were 31x more likely to be infected with COVID than boosted...
  • Lou_BC: Context is everything: “Despite fewer than 13 percent of adult Massachusetts residents being completely...
  • Oberkanone: Insufficient demand to build in North America. Importing Zephyr allows enhanced lineup at low volume...

New Car Research

Get a Free Dealer Quote

Who We Are

  • Adam Tonge
  • Bozi Tatarevic
  • Corey Lewis
  • Jo Borras
  • Mark Baruth
  • Ronnie Schreiber