It's funny how most amateur boatbuilders (and probably even professionals) will spend thousands of dollars on good tools over the course of their hobby or career but shy away from vacuum bagging because they don't want to spend $350 on a decent vacuum pump. I suspect it has more to do with the complexity of vacuum bagging than the cost. True, vacuum bagging adds a bit to the cost of your boat's materials, but the improvements in structure and finish are well worthwhile. If you wanted a cheaply-built boat you would have bought one from a mass-producer anyway. For me, the simple reduction in the amount of sanding makes vacuum bagging worth it. Most home builders wouldn't even consider building a boat from sprayed chopper-gun fiberglass, so why not take the next logical step and try to attain aerospace quality?

I have played with many different forms of vacuum bags. Most of my efforts are concentrated on getting a good seal and not wasting a lot of disposable materials.

The Setup

I use a Gast rotary vane vacuum pump connected through a 25-foot air hose to a 5-gallon air tank. The tank has a vacuum gauge and switch and connects to the vacuum hoses, which connect through a vacuum port to the bag. The long hose is helpful in getting the loud pump as far from the bag as possible so you can hear the tiny leaks. The tank helps to avoid cycling of the pump when the bag reaches the desired vacuum, and serves as a resin trap if any resin gets sucked through the hose. I realize that most designers don't recommend bagging more often because it's difficult to get it right. In fact the only designer I know of who incorporates vacuum bagging into his plans is Derek Kelsall and he's got a patented system for bagging hull panels with a shop-vac directly on a waxed table. It's theoretically much easier to bag on a flat surface. The biggest problem is getting a good seal at the edge of the bag and this is definitely easier on when you can press down on the table.

Sources for the equipment are:

Vacuum pump: Cole Parmer.
Switch and gauge: ACP, Fiberlay, Fiberglasssupply.
Air tank and hose: Sears.
Fittings and connectors: brass, from local plumber's supply, with teflon tape.
Vacuum port: Fiberlay's through-bag port works best.

The switch is adjustable from 0-30" Hg. The check valve prevents air from leaking into the system through the fittings or the pump, particularly after you've achieved your desired vacuum and the pump shuts off. The relief valve lets you rapidly release the vacuum and it's useful for establishing just a few inches of vacuum initially so that you can arrange the bag to make sure there are no wrinkles at critical spots (like the edge of the laminate) and no voids where the bag is sealing a bubble of air.

Make sure that the gauge and switch are tee'd off the suction to the side, so that any resin that may get sucked into the line will drop into the tank instead of the pump, gauge, or switch. You really should never even come close to getting resin in the lines because you should always use many layers of breather/bleeder felt under the vacuum port, and the port should never be directly over the laminate.

I use a Tanita electric scale, accurate to 5 grams, with a capacity of 2kg. I doubt that I'll ever want to mix 2kg of resin at a time, but I doubted that I would ever use fast hardener when I started this hobby and now even fast hardener is usually too slow, so I may yet start using large batches of resin.

The scale is more accurate the more resin you mix, which is fine because my largest batches are for laminating, where accuracy is critical, and my small batches are for filling and fairing, where accuracy is less critical because strength is less critical.

I will never return to hand pumps. They just don't work well enough, and tend to spit air unless you pump unbelievably slowly (1 stroke/ounce per 30 seconds). I've tried everything, from heating the resin to holding my finger over the pump so it can't suck in air, and I've used many different pumps. Buy the scale.

I generally mix about 25% extra resin. My current resin of choice is from Applied Poleramic. I weigh the fabric, then pour at least the same weight of resin into my mixing cup, and consult a chart on the wall for the amount of hardener to add. Most resin suppliers will tell you appropriate ratios by weight, but if not you can calculate by pouring identical volumes of resin and hardener and weighing each, then adjusting the mix ratio by the respective weights.

When cutting out a panel that's going to be bagged, always make the panel at least 1/4" larger than it's final size in each dimension. When you apply the fabric to the panel, it's impossible to maintain an exact shape, since every stroke of the squeegee shifts the fabric around. So you have to cut your fabric slightly larger than the shape of the panel. But the vacuum bag forces the ends of the fabric to wrap down the edge of the panel, so if your panel is 1/2" thick you end up with a 1/2" thick mass of plastic and glass at the edge of the panel. It's much easier to trim the panel by cutting through the laminated section than by cutting through this lump of solid resin, so make the panel larger than you need and trim it later.

Also, every time you handle the fabric (particularly with carbon or bidirectionals), you're going to lose one or two tows from the edge, which makes your piece smaller each time you pick it up. Taping the edge with masking tape prevents this, but you can't get resin under the masking tape so it can't be part of your laminate.

Materials

I've tried using bubble wrap as a substitute for breather/bleeder felt. With bubble-side down it effectively absorbs excess resin in the spaces between the bubbles but can only be used once this way before clogging up. Cleaning is possible but messy. With bubble-side up it wont' absorb excess resin and you risk voids from trapped air. Either way it can only be used once or twice under high vacuum because some bubbles lose their air after the first time and result in voids, or at least a really bumpy surface. Stick with real bleeder/breather felt. ACP has some good felt made from recycled material.

Very light CSM can also be used as a breather/bleeder, but it's messer to handle than felt because it leaves glass stalks everywhere.

Some 4 mil poly dropcloths work well for the first lamination, but develop pinholes from handling and can only be used a few times. I say some because many are made from recycled material and have many pinholes even when new. The "milkier" the bag, the worse it's going to leak. Completely clear poly cloth is best. Any thinner than 4mil and it will immediately be punctured, especially by raw carbon fibers. Any thicker and the bag will crimp and leak in many places. Fold over the edges and hold with clothespins before sealing with the bag sealer. Don't use poly bags for anything after the first lamination because sharp protruding edges from the last laminate will puncture the bag. It's easy to seal edge leaks with the bag sealer by just crimping the bag around the leak. With a new bag, it's possible to get 26" Hg (the max for my pump) that drops down to 20" in over an hour (after the pump is turned off), on a 16-square-foot bag being used for the first time.

Genuine nylon bagging film is much tougher but can't be melted shut with the bag sealer. It only lasts for three or four laminates anyway.

My best bag so far is made from 6 mil polyethylene tubing, available from Fiberglass Supply. I clamp each end of the sleeve between a strip of hardwood and a 2" square strip with foam weatherstrip attached to it. This works so well that I have come back to the shop the next day to find the bag still under a few inches of vacuum, after the pump has been off for hours. The limitation is that it's only useful for panels under 30" wide, and about 6 feet long. The easiest way to get a long panel into the bag is to use a piece of stryfoam insulation (covered in plastic) as a base to laminate on, then slip the insulation into the bag, pizza-oven style.

You have your choice of using perforated release film, peel ply, or both. Don't use release film if you will be laminating further, because it leaves a very smooth surface, which needs to be sanded before laminating. Using peel ply over release film seems unnecessary. For thick laminates that are going to be oozing a lot of resin into the breather, you might want to use release film between the peel ply and breather, to make it easier to peel off the breather.

Sealing the Bag

The most common technique is to use either dedicated vacuum-bagging mastic or 3M window glazing in tape form. Either works fairly well, particularly on odd-shaped bags, but both are expensive, since they generally can't be reused without making a complete mess. The don't work for me on bags whose surface area is more than a few square feet, or on bags which you're attempting to attach to uneven surfaces. The most important technique to remember is to lay the bag on a flat table and attach the tape to the edges of the bag, instead of attaching the tape to the laminate and the bag to the tape. With the tape on the bag, it's easy to seal creases or pleats that inevitably show up because it's impossible to lay out the bag perfectly flat and perfectly square.

A 6" electric bag sealer works well to seal poly bags. The bag can be reused once or twice if you bag your largest part first and cut off the sealed edge.

I've tried to make my own bagging table by building a wooden frame, putting a strip of "Weatherproof Boat and RV" weatherstrip around the bottom edge, and clamping it to a folding table, with two sheets of film in between. This has been moderately successful, but if there are any leaks they are hard to stop since you can't lift the laminate off the table and you can only put so many clamps around the edge.

If your laminate can't easily be slid into a bag (like a pizza into the oven), the above methods work best because you can do the lamination on the bottom layer of the bag, drape the top layer over, and seal the edge, either with the frame or the sealer. If your laminate can be slid into a bag or tube, you can generally achieve much higher vacuum because the leaking edge is much smaller. I have four strips of 2x2" wood with weatherstrip on one edge which I use to clamp each end of a bagging tube and this enables much more reliable high-vacuum generation, since the potential leaking seam is much smaller.

Next, I tried to recreate a setup I saw in an old issue of Fine Woodworking magazine. I've glued together two vinyl shower curtains and I use a piece of CPVC (plastic pipe) to seal the bag. You wrap the open end of the bag around the pipe and clamp another, larger diameter pipe around the first. So far the larger pipe isn't flexible enough to snap around the smaller pipe so I'm still pondering this one. Folding and clamping the end of the bag results in too much leakage.

Another useful tip from the magazine is using clamps, hung from the ceiling, to hold the bag open.

Eventually I settled on a small-diameter PVC pipe for the inside, and a piece of foam pipe insulation for the outside.  This works best for me and results in bags with zero leakage most of the time.

Unfortunately the vinyl shower curtain doesn't seem to be very airtight, so I'm back to nylon and poly bagging sleeves, which generally last about 4 laminations.

Here are some photos from the laminating process for one of the CMM central panels.

I tried a brief experiment in resin infusion. I found some landscaping material at the hardware store that looked like a good candidate for the type of material that would allow resin to flow through. My resin was a bit too viscous but it worked reasonably well for a square-foot panel.

In the picture, the hose on the left is 1/4" garden watering hose, and leads into a cup full of red-dyed resin. Inside the bag, it goes into a T-connection attached to 1/4" garden soaker hose. The vacuum connection is on the opposite side of the bag. You can see the resin bleeding through the fabric. I conclude that you would have to have a lot of practice to figure out where to put the resin tubes and the vacuum ports, which is probably why variations on this method have been patented since the 1950's with the "Marco Method."