Adam Grace [APU FOREMAN]



MATRIX: How did you get into the film industry?

ADAM: I’d seen Star Wars when I was still at high school and I thought that was fantastic. I was doing model making as a hobby, but when I saw that film I thought I’d love to get into that, so once I left school I kept that going as a hobby and started making contacts in the TV commercial and film worlds. My first job was at an engineering model making firm and that was a fantastic grounding for doing work like this – working from scaled drawings, working with different materials, learning molding – and then eventually got into TV commercial and film work from there.

MATRIX: What projects have you been involved in?

ADAM: I worked on the first MATRIX, where I was involved with the replica Bell helicopter that appeared in that film, and a small amount of finishing work on the quarter scale helicopter. I was also involved with some of the set dressing pieces and components for the Nebuchadnezzar hovercraft. THE MATRIX was probably one of the first popular films I worked on; I had worked on other films before, some sci-fi, that didn’t do quite as well. I worked on The Time Guardian [1987] and was involved with some miniature work for 20,000 Leagues Under The Sea [1997]. There were two productions of that; this one had Michael Caine in it and wasn’t for theatrical release.

I worked on Mission: Impossible II and I did some work on Total Recall when it was going to be shot here [in Australia in 1987] – we got within four weeks of filming – it was a Dino De Laurentiis production – and the whole thing got canned, and all the miniatures we’d worked on got packed up in containers and years later it was remade in the States totally starting from scratch. It was slated for Bruce Beresford to direct with Patrick Swayze in the lead, and I hadn’t read a script but the design was different. For the Mars scenes that we were working on, in some ways the design was a bit more ornate, almost Victorian if you like.

Then a lot of those models that were packed up were subsequently bought by a company we worked with to produce the Ultraman Japanese TV series, where we looked after the miniature effects. There was a scale model of a New York street set so all those modern-day buildings were reconfigured for a downtown city being attacked by giant monsters. And some of the sci-fi-like components were turned into futuristic weapons for defense against these monsters. So nothing was wasted, everything was eventually used and seen on a TV screen.


MATRIX: What did you do on the Nebuchadnezzar for the original film?

ADAM: On the Neb I was involved with the control console off to the side of the Operator’s chair that had a keyboard and a disk drive with a little tray with the disks on it. In the actual cockpit of the Neb there was a central console and on either side of the console were two steering controls, which you probably never see, but I made those as well as a strange device that was supposed to be a hologram projector, which again I don’t think you see. I think they had something for light to come from, but off to the side there was an arm with a dish on it that was supposed to be projecting it, and that was then (I think) suspended over a magnetic field, or whatever the technology was for that. When they came to rebuild the Neb for the second film all that was gone, so I saw guys in the workshop remaking the bits that I’d made for the first film.

I also worked on the Neb Infirmary scene where Keanu has got hundreds of little needles in his body. My job was to silver solder spring steel “posts” [of the needles] to little punched disks. I silver soldered hundreds of those together that then went off to makeup effects for them to insert in their prosthetic makeup and then place on set. I helped make what they called the “robot doc” above the acupuncture bed where all the wires came out of. And I did a little bit of work on the “grappling droid” that plucks Neo from the sewer when he is released from the pod. I was involved with some molding of the components and then it was engineered after that – a lot of engineering work went on after my job was done.

MATRIX: Did it have a different feel the second time around?

ADAM: Yes, it does have a different feel. On the first one, even though Keanu had been in a few films, people weren’t sure whether it was going to work or not because it was different. It is a different feel now, knowing the success the first one had. There was nothing of the scale of the APU on the first one… although I was involved with the helicopter on the first one and that was five meters long [16.4 feet]. It was a full size helicopter and weighed three tons, so that is a similar size object. But there was nowhere near the amount of components and information needed to produce the APU. With the helicopter we were working from a known object, which made life a lot easier.

MATRIX: How were you involved with the helicopter?

ADAM: There were two aspects of that: there was a full size helicopter that matched one they were using for real flying sequences – Neo on the side shooting into the side of the building, and then for the purely crashing scene there was a quarter scale miniature version. Primarily I worked on the full size helicopter, while the quarter scale one was happening at the same time, so I virtually came in at the end of that and just helped paint it. The chap in our workshop that did a lot of the pattern work in making the quarter scale one is also working on the sequels, Gary Cameron [Prop Manufacture Leading Hand].

MATRIX: Why not get a real helicopter rather than build one?

ADAM: We were asked that a number of times and in Australia there isn’t a graveyard of helicopters like you can have in America. Also, a helicopter that is made for a film is made specifically for filming – that is it has areas that are what we call ‘wild’ that are removable. We could quickly go in and remove a tail section, or a wall, or a seat, or the whole windscreen, which the prop helicopter is designed to do, whereas a real one isn’t and you would be working with something that is virtually all steel. With the model, if the Grips, Camera, or Lighting wanted to come in and punch a hole in it and put in a light or a camera and clamp it onto the frame inside, we could just cut through our fiberglass skin and do that in no time. We can cut metal quickly as well, but fiberglass is an easier material to work with.

MATRIX: Do you remember how long it took to build that?

ADAM: There were parts that were started before I came on – probably a month or two of pattern work done beforehand – and then I came on for three months, so it was probably a five month project from start to finish. We were given a wrecked helicopter body from a collector that was the same model as the real one they were using for flying, it was just from the nose to where the tail section joins the body. From the collector’s damaged helicopter we took fiberglass molds and then reproduced shells out that went onto a steel frame the steel construction crew had welded together, and then returned the helicopter. Then any parts that were missing we made patterns for in our shop, took molds off those and then fitted those as well. We photographed extensively the helicopter that was used for flying, and then we matched the interiors of that.

Anywhere the panels were missing we’d infill and in the process of molding, even if it was brand spanking new, we had to plasticine over any seams to make the area we were going to mold impervious to the resin that would’ve soaked in behind and caused the mold to lock onto the shell. We had to be careful – we had to make sure that we had the proper release agents on it, and that any infilling with tape or plasticine was removed afterwards.

MATRIX: Why was it better to use the collector’s wreck than the actual helicopter?

ADAM: To have the actual one out of service while we molded it would have cost a fortune, whereas we were able to utilize the collector’s one and not tie up a real helicopter for weeks while we molded it.

MATRIX: Where did the flying helicopter come from?

ADAM: We used an existing one in Australia, and it was painted black for the film and then had to be repainted back to its original livery again as part of the deal. The helicopter that was hired was a freight helicopter; it was white and green with black stripes.

MATRIX: How do you ensure that a full size and miniature object match in texture and feeling and weight?

ADAM: In relation to the weight it’s a bit different for the one in THE MATRIX because it was mounted on a rig that caused the helicopter to have a predetermined path. Weight as far as look and feel wasn’t a factor for the quarter scale helicopter. Things on the miniature that we had to take into account were that the rotor blades had to be able to crumple and break as they hit the quarter scale building, and there were panels on the quarter scale helicopter that had to crush in as they hit the building. So attention was paid to how to replicate the skin of a real helicopter and its underframe in miniature. We had to duplicate it for that effect to take place with the miniature helicopter crashing into the building.

As far as painting, because the real helicopter was finished in a satin or matte black it pretty well dictated the finish on ours, although we probably had more weathering on ours. Generally, helicopters in domestic use are fairly well kept and are very clean, but in miniatures you tend to – even for clean things – accentuate rust marks or a bit of oil stain. So by the time it’s filmed and it’s traveling through the screen those marks will read, and the eye will pick out that there is some detail. You tend to accentuate details like that to help miniatures read on screen.

MATRIX: The miniature had to be on screen for a good amount of time because of the crash.

ADAM: Yes, often the size of the miniature is dictated by that very fact, in that it can’t be too small otherwise when it gets destroyed the broken parts would just be too out-of-scale and it would give the size away. It can’t be too big otherwise you might as well end up using the real one. All the rigging that would be involved with moving a real helicopter and having a building destroyed is too much. There’s a trade off between what’s going to be manageable as far as rigging and effects and setup, and what is still going to be big enough to read and behave properly.

MATRIX: Do you remember the crew members involved on the helicopter project?

ADAM: On the quarter scale Gary Cameron, Aaron Crothers and Ronald Rametta did the principal work, and I’m pretty sure Jo Thompson got involved with some mold making, and Norman McGeoch worked on some of the mechanics in the quarter scale, and Tom Davies our Miniatures and Models Supervisor was involved, obviously. They were the core on the miniature team, and then there was a group of people doing the quarter scale building façade.

As far as the full scale helicopter, I worked a lot with John Murch and Mark Powell – the three of us were involved with most of the mold making and the assembly. Simon Parker and Peter Milton were involved with some of the initial mold work, as well as Dion Horstmans and Kris Barron.

MATRIX: What was it like to see the final sequence?

ADAM: Fantastic. I remember they had footage on tape and they put it on a TV for us in the Art Department, and those of us who’d been involved with it had a look and it was fantastic. And that was before the visual effects had been added, that was just the raw footage from (I think) six cameras in various places; it was all the takes from the six cameras. We still even then didn’t realize the type of film we were working on, but it was exciting to see this thing swing in and hit the building, see the ripple or shock wave, and then the building implode and explode.

MATRIX: Knowing that there’s no such thing as a small action sequence in Larry and Andy’s films, does that affect the way you work?

ADAM: I try and put the same effort into everything I work on, but there is something about knowing that working on something like the APU – anyone that sees it gets a big kick out of it, and I feel proud to be involved with it. I’ve also heard about other sequences when I’ve talked to other people on the crew, and I haven’t seen any footage yet but I know it is going to look amazing. It helps being part of something that you know is going to blow peoples’ minds. I don’t know if it makes it any easier, but it certainly makes it enjoyable.


MATRIX: How long have you been working with the APU; the Armored Personnel Unit?

ADAM: I’ve been working on the APU for a year. We had a break at Christmas [2001] and Easter [2002] and a quiet period around the end of November [2001], but I’ve been on APU and nothing else for a year.

MATRIX: When you first got the job did you realize how epic the piece you were creating was?

ADAM: I realized it was a big project. The Art Department prepared some full size outline drawings of the APU plus some renderings, so I knew the physical size of this thing and I knew it was going to be a long job, but I didn’t think it would take a year before it ended up on set.

MATRIX: How was the size determined?

ADAM: There would have been a design choice done with the Brothers [Larry & Andy Wachowski] and Geof Darrow [Conceptual Designer] when doing the initial designs and sketches. I’d seen some drawings of Geof Darrow’s, and they were indicative of a size just by his drawings.

MATRIX: What has been your role on the APU crew?

ADAM: My part was to coordinate a team of model and prop makers, and our job was to produce the patterns from the supplied drawings to duplicate shells and parts that would fit over the mechanical understructure that Martin’s [Crowther, APU Engineering Foreman] team was looking after. These patterns were based on what the object had to do, how it had to perform, how we might need to access certain parts of it, and whether they were going to be fiberglass shells out of molds or vac-formed components and patterns.

We worked very closely off Sergei’s [Chadiloff, CAD Computer Modeler] drawings. We followed them as close as we could unless there was a change required for a material availability to streamline the process of manufacture. For instance, in vac-form you can’t have negative draft on objects, so we would introduce a few degrees draft to enable the vac-form shell to release off a pattern. Vac-forming is a process of clamping a sheet of rigid plastic in a frame, which is heated up until it’s pliable and then it’s brought down or draped over an object, and with a vacuum applied underneath it would draw the air out between the sheet and the pattern, and the plastic then would take on the form of the pattern. Once it has cooled you can take it off, trim it, and you have a ready-to-go rigid shell copy of the object.

MATRIX: Please explain what “negative draft” means.

ADAM: The sides on your pattern can’t have an undercut and really can’t be ninety degrees (even though we did fudge a lot of things off vertical patterns) otherwise the shell will lock on to the pattern. So it has to be wider at the bottom than it is at the top, if you like. That applied to virtually all the patterns because most of the molds were fiberglass molds – we had to get a rigid mold off a rigid pattern and then cast a rigid shell out of it. The only molds that then went to, say, a silicon molding process were those we couldn’t achieve easily by altering the pattern without draft. In that case we would use a silicon that enables you to stretch off a pattern and cast complex shapes out of it.

To create the silicon molds you again start with a pattern and then it is encapsulated or covered in a silicon molding compound, which is a two part component – a liquid rubber and a catalyst. Once it has catalyzed it would go solid, but it has elasticity, so you’re able to stretch the mold off the pattern. That means you’re able to do complex shapes with undercuts or a lot of surface detail and be able to stretch it off the pattern and have it reproduced.

MATRIX: Why couldn’t you just have two halves and mate them?

ADAM: Typically most of the patterns and the corresponding molds and shells were produced in halves and joined around the armature. Molding decisions were made based on size and cost; silicon is not a cheap material, whereas a rigid mold like fiberglass is a cheaper material. If the object was too large then we would go with a rigid mold with parting lines. The fiberglass molds were generally multi-part molds, so the complex object was encased in a many part mold – two, three, four. We tried to keep them as minimal as possible, but for some of the shapes we had to go with a four or five part mold just to enable us to unbolt the mold shells off to release the pattern and the cast object. Typically there were three chaps in the mold department doing this.

MATRIX: Can you illustrate what you have just explained using an APU leg.

ADAM: The pattern for the shell would have been carved out of jelluton, which is a pattern-making timber, and then a fiberglass mold would’ve been taken off and split down the center, so we had two halves of the pattern. Then we would lay up fiberglass in that fiberglass mold and reproduce the fiberglass shell, and we’d have two halves that would join around the metal armature. The casting of the cap is just a vac-form because it’s such a simple shape and we can pop those out in minutes, whereas a molding lay up process like for the leg typically requires days.

MATRIX: How many parts were you able to pull from the same mold throughout?

ADAM: As many as required. One of the changes we made to Sergei’s design was a small change that enabled the part to be handed – that means it could serve both left and right sides – and that saved us time in making only one pattern and one mold, and just multiple castings. There are parts that are left- and right-handed, for which we had to make left and right patterns and left and right molds. The arm lent itself to being even handed but the leg, due to the nature of it, we really couldn’t make too much of it even handed, as far as, say, the thigh and the shin. The toes could be duplicated on either side, but we had to do left- and right-handed ankle pieces.

MATRIX: The scale, size and sheer weight involved – was that at all daunting?

ADAM: No, at the beginning I thought to myself if I try and think of the whole thing all in one go, it’ll seem too much, but I knew it was going to be a challenge. So even though we were discussing how the whole thing may go together and we were estimating weights, when it came down to manufacturing the parts we were just looking at subassemblies. What was helpful in some respects was that we only had drawings for subassemblies. When we started at the toes they were still designing the torso and the top half, so we couldn’t go too far down the track thinking about how we were going to make those bits. Certainly by the time we got a third of the way up we had a system going where we would generally know what size of object we were going to do as a fiberglass shell and not a vac-form.

MATRIX: The APU is almost “macro modeling.”

ADAM: It is, it’s the same skills if you like. We still have to finish patterns well enough to be able to have molds release off them cleanly, so there are the finishing skills, and the ability to read engineering drawings is required. It’s just that instead of the whole model fitting on the bench, only the one part you’re working on can fit on the bench.

I’ve been privileged to see some animation tests and they look fantastic – we’ve made a static prop and when these guys make it move with their CG, even the wire frame stuff looks fantastic. I can hardly wait to see the APUs stomping down the Zion Catwalk.


MATRIX: How heavy is this the APU?

ADAM: It’s about two and three quarter tons – probably just shy of three ton [6,063 to 6,614 pounds]. We weren’t given any weight restrictions, but our initial estimates were around two tons [4,409 pounds] and it grew from that.

When we were initially looking at it we figured that regardless of what shells we dress on the outside, it is going to have to stand up to being used on the film. Even though it isn’t self supporting, it had to be structurally capable of being moved around and supporting its own weight as far as handling goes. There was no point in putting a lightweight armature inside so that the moment we moved an arm or something, we tore the pivot points off.

MATRIX: There’s going to be a lot of CG work done with the APU, but putting that aside what physical effects are there going to be?

ADAM: It’s jointed so it’s capable of assisted movement. For instance, if we want to see the initiation of an arm moving we can do that by having a cable rigged to the ceiling, and when the cable is wound in the arm will move. Our brief was that an initial movement was required, and from there on CG will take over. As far as physical things we have the console control arm that will retract into position once the Operator operates the levers. There is a pneumatic shaking rig provided as well. Our department has built small units in the shoulders, but the one that’s probably going to be predominately used is a large one built by the Special Effects Department to make virtually the whole APU shake to simulate the firing of the guns. And there are some pneumatic rams in the top pistons of the shoulders that are capable of giving some extra movement – if the shoulders were to move we can activate those just to make it look a bit busier. We also have ammo box holders that you see on either side of the legs that have pneumatic operated doors and an ejection tray.

MATRIX: What was it like the first time you saw it put together?

ADAM: When we were putting the leg together we got the left leg up and it was standing there and I was looking at it thinking… this is just its leg and it’s taller than us! All the way along people had commented on whether we’d over-engineered the steel because we were looking at some heavy steel sections for the pivot points and the joints, because there’s a lot of load bearing down. I have to say that it was good to go down that path because Martin made the right decisions – anything less and you wouldn’t be able to stand beside it comfortably.

It has to survive six weeks of being placed on its back and its front and kneeling, and people climbing in and out around it. We had some of the components made in cast aluminum so we could save time, whereby the pattern that was made went straight to the casters and came back as cast aluminum pieces and were ready to machine. We didn’t have to make shells and armatures inside them.

MATRIX: How does the APU Operator get into his chair?

ADAM: When it’s in the kneeling position the whole unit drops down and he’s able to climb up the leg onto the knee and then step into the foot plate then into the seat.

MATRIX: Is there anything interactive in the chair for the Operator?

ADAM: Right now I’m sitting in the Operator’s chair about two and a half meters [8.2 feet] off the floor in the front of the APU. When the actor climbs in he’ll flick a control lever and that will cause the console to come up in the closed position. He’s got a few switches and knobs to keep himself busy should the Directors call for that, as well as a couple of lights. His main action is to have his hands in the control arms, which are jointed and telescoping, and as far as we’re aware the Directors will say where he should shoot, and he’ll point his arms and pull the trigger.

Of course they don’t want an actor in there waving their arms wildly because that’s going to be costly CG in post. We’ve also got control in the footplates as well as pedal control. They’re still under debate about whether the Operator actually walks the APU by stomping along, or it’s just forward and reverse done by the pedals. The decision was made to not put a canopy over the Operator to contrast the huge firepower of the APU with such a vulnerable human out front.

MATRIX: What sort of tests did you do for the shaking of the APU?

ADAM: Over at Special Effects we tested the shaking rig, which had the whole APU suspended by one cable. We had the shaking rig going, and I was sitting in the carriage [Operator’s chair] mimicking any movement that the guns were doing. It was shaking around so much that the first time I did it I wondered what I had gotten myself into… but I settled in after that. The test was mainly to see how the APU and the actor would hold up. With the special effects rig going everything shakes and wobbles, and the initial feeling when you’re sitting there is that there’s nothing underneath you, and you hope the whole thing doesn’t fall over – which of course it won’t.

MATRIX: For different scenes the APU needs to be moved around; how did you simplify that on a three ton object?

ADAM: To facilitate when we break it down for moving, or for taking parts off the APU for shots where they’re going to do a CG part, all the joints have a basic pin system in them so we can access those relatively quickly. We have dressed all the joints to the design, but made the dressing caps pop off easily to access the joints. So if you were going to remove the gun from the wrist you would just remove the cap, undo the parts, drop them all off, take the pin out, and then we can remove that gun easily. Once we redress it we just pop the caps back on – they’re all labeled. With the look of the APU and how it’s designed, virtually all the joints have some type of a cap that’s either a friction fit or is clicked into position.

MATRIX: What are the springs made from?

ADAM: The springs are made out of coiled copper except for the larger ones, which are coiled copper that has been covered in an insulation hosing. The springs between the carriage and the torso boards – there are two boards that are parallel with each other on the front of the torso – are coiled airline hoses dyed black. Because those parts actually move, that allows those to flex in there, whereas even though the other springs look real, those parts don’t actually move. The rest of the hoses are all flexible and we’re able to disassemble them quickly, so when we remove parts we can take it apart quickly and put it back together again.

MATRIX: How much planning went into the creation of the APU?

ADAM: Oh each part that has been made on the APU, there were discussions with Martin and myself as to which armature will go inside it, and whether it will have to be removable when we take that section of the APU apart. We then decided on how we would make the pattern, and what type of mold and what type of material it would be made in. We were constantly thinking about what it had to do, and how it had to behave.

Even though it is now made, there’s a new phase of work that has just begun where I now have to make it work with a shooting crew and the demands that all that entails – and make sure we don’t hold up production.

MATRIX: What is your role now that the APU will spend its days on set?

ADAM: My role is to help prepare the APU for whatever setup it’s required for. Special Effects will be in charge of moving it from place to place, and my job is to dress or prepare the APU for how it’s supposed to look on set for that particular scene. It might mean disconnecting cables or hoses, it might mean clipping on or attaching panels that redress it into a different APU, and I will have to remove parts and help assist in the breakdown when we transport it from set to set.

MATRIX: Do you know how many different locations is it going to be shot in?

ADAM: At this stage it’s Stage 1 at Fox Studios Australia, and then on a stage at our temporary studio in Matraville.

MATRIX: What kinds of concerns do you have as regards moving and positioning the APU?

ADAM: We have done some tests, and when we start stringing it up and tipping it over it does put stress on the APU joints in the internal structure. So when we’re turning it on its side and placing it on its back – the side is probably the hardest position for it – we have to be careful not to let the crane cable out too fast. I’ll also make sure that I’m keeping an eye on any of the joints or the parts that they’re not crushing anything. You have to get those parts in the right position so they’re not going to damage anything. But if we are careful and do as we’ve rehearsed then we shouldn’t have any problems.

MATRIX: When it is standing on set, is it attached to a crane?

ADAM: No, it was brought in on a crane and attached to an overhead truss with a cable, but there will be setups especially when it’s laid on its side back or front where we’ll have to have a crane to do that… in fact two cranes to enable it to do that.

MATRIX: Thank you very much Adam.

Interview by REDPILL
June 2002