FIRST GFRC PROJECT WITHOUT SPECIAL EQUIPMENT I NTRO DU CTI O N After scouring this forum for information and learning as much as I could about GFRC (see Concrete 101) I was ready to get my hands dirty. The plan was to start simple with a small easy piece and work my way to the more complex. Unfortunately due to circumstances beyond my control I had to start with complexity. I have some experience with concrete but none with GFRC. I am doing the work myself for the fun and satisfaction of it and will never be a pro. This situation doesn’t justify buying expensive equipment that might make the process a little easier or more successful, but it sounded like it was possible achieve reasonable results anyhow. Below is my experience with GFRC and the equipment I had in my garage (with a few exceptions as noted). T HE P RO J ECT My first project was a vanity top with integral sink. Many of the parameters were already set by the odd configuration of the vanity. 15” long by about 60” wide. Asymmetry was important due to our faucet set (see photo). I bought Gore Design’s Teton sink mold as the key element of the design (amazing design and beautiful product…thanks Brandon!) and added a flower/toothbrush holder for fun. I wanted natural concrete with a crisp, clean, minimalist look. Since I don’t have grinding equipment to expose aggregate I opted to go for a cream finish. Easy, right? M ATERI ALS I went with the Blue ZFRC system due to the reviews on this forum, the cost/performance ratio, and the tweakability of it. Although I would have loved to try a from-scratch approach to mix design there are just too many add mixes and combinations to test so I thought the better of it and started with a tried and true system. (Good choice!) My goal was to produce a smooth, even finish with almost no activity (similar to a quartz countertop). I figured that if I could do this I would have good control of the craft and would be able to experiment with adding character as time went on. It seemed like good mixing and the spray technique would make this relatively easy. T OO LS I used my rusty but trusty hammer drill with a small helix mixer in a 5 gallon bucket. It’s a large corded drill and had plenty of power for the job. I didn’t try anything smaller. I didn’t have a hopper gun and opted for the $19 Harbor Freight model that seemed like a direct copy of some of the better models discussed on this forum. It got good reviews so I thought I would give it a try. I used my 5.2 cfm (4 gallon) compressor which seemed to be able to keep up without a problem. On the finishing side I wanted to avoid buying diamond pads by going with a cream finish. My plan was to use Scotchbrite pads on my angle grinder to remove any water marks and polish the surface. More on that later…
F O RM S I made melamine forms with the Gore sink and slot drain system. The form making was quite straightforward thanks to the great instructions on the Gore web site. I actually found a PVC board remnant at my local plastic dealer for the slot drain as suggested. I almost substituted melamine I had on hand, but since I didn’t know why the Gore instructions suggested PVC when melamine is so cheap and available I didn’t want to take any chances. This was probably the single best minor decision I made on the project. It wasn’t clear to me how the pieces would separate as I demolded. It still isn’t. I figured that would sort itself out. It did. Sort of… I read Ching’s book and used his form making techniques. I forgot the plasticene to keep the concrete out of the screw heads so I just ran a strip of painter’s tape across them all. That worked fine. The only real issue I had was that some of the screws telegraphed through to the surface by creating a very slight crease/bulge in the melamine in the receiving piece. The only way we resolved this was by pre-drilling which was incredibly time consuming. We were using 1-5/8” drywall screws that I had on hand. Suggestions? M I X D ESI G N I used the Blue ZFRC/Gore mix design. They were almost the same so I can’t remember which one I followed more closely. I didn’t use any fiber in the face coat, and I used locally sourced gray cement since I wanted a traditional gray color. I used the superplasticizer I already had on hand, and also used basalt fiber for the backer coat because it looked cool and I had never seen it before. M I XI NG I had no mixing problems at all. Just followed the instructions. S PRAYI NG Since I had no experience spraying concrete I thought I better do some tests. The first attempt was pretty disastrous. The concrete just kind of spit/dribbled out of the nozzle. Apparently the pressure was too high. Seems counterintuitive but I had it at about 100 PSI which was higher than the gun’s max. I had forgotten to reset it, but I would have guessed that it would have atomized the concrete too much and perhaps caused the sand to bounce out. The opposite occurred. Weird. Anyhow, after lowering the pressure to 50 psi and switching to the medium nozzle (from large) I got better results. After some experimenting I settled on 60 PSI with the large nozzle and fairly thick concrete. Probably more like oatmeal than a milkshake. This seemed to hang on the verticals better than the thinner stuff. After two perfect tests in a row I was ready to move on to the real thing! T HE FACE C O AT The spraying started beautifully, but quickly turned to horror. The nozzle went back to the spitting and dribbling of the first test batch. It seems as if I didn’t control for the time element in my tests and the concrete was either setting up in the hopper, separating, or otherwise clogging the nozzle. I quickly dumped the mix back into the bucket, remixed it, cleaned the gun, and resumed spraying. Of course now there was a seam, as well as all kinds of dry, sand-like concrete particles in the unsprayed portions of the form to deal with. I tried to whisk them away as well as
I could with a chip brush, but it was very difficult to get them out of the corners and I could really just move them around. I tried to brush them on to the concrete that was already sprayed with limited success. I should have had a shop vac ready. This clogging happened twice more and I had to temper the mix with water the final time to get it thin enough to spray. I was using the Buddy Rhodes brush technique between coats and that went fairly well except that I tore the surface occasionally. I found that I could place the brush on the surface and vibrate it to repair the tear. The face coat seemed to liquefy when vibrated just like any concrete. This all happened very quickly, probably more quickly than it has taken me to describe it. I would say the spraying lasted 20 minutes start to finish, including the brushing and nozzle cleaning. I was moving fast as I feared the concrete would not mend together each time I stopped. In hindsight the thick concrete wasn’t such a good idea. I will have to do more experiments, but the milkshake-like consistency that Buddy Rhodes used probably wouldn’t clog like mine. I hope that I will be able to get a full coat on before having to stop and clean the gun. T HE B ACK C O AT I mixed the back coat in two batches. This was a good call even though the piece wasn’t very big. I was inexperienced so moving relatively slowly. I had also mixed the first batch quite thick. Probably cookie dough like or even thicker. It started out fine, but it took close to an hour to get it down due to the complex shape of the form. I had to temper it towards the end as well. The back coat was so much easier to work after the tempering that I mixed the second batch a bit thinner. More like pizza or bread dough than cookie dough. This was much easier and quicker to work with. The second batch went on in half the time. C URI NG After cleaning my tools I misted the surface with a spray bottle and covered it with painter’s plastic. Then I placed a doubled-up electric blanket I bought on closeout from Target over the whole thing. It was tent-like due to the drain assembly sticking up in the middle, but I figured this was good as I didn’t want coil marks to telegraph through to the concrete surface. I left the blanket on high overnight but the surface seemed to have dried out in the morning. I sprayed the surface quite aggressively until there was water standing on the back surface. Then I recovered in the plastic and blanket and let it continue to cure until 48 hours after I had poured. D EM O LDI NG The melamine came off quite easily but the sink form stayed in place. Unfortunately I had done a very good job attaching it to the slot drains and they weren’t budging. I was trying to tap them out with a hammer from the back as instructed, but no joy. The sink itself was clearly not adhering as I could flex it up, but the slots were keeping it in.
Violence! I just kept hitting it. I went from a hammer to a sledge hammer. Finally it seemed like it had moved 1/32nd inch. Then 1/16th. It was moving! It took over 2 hours and probably 2000 blows, but it slowly made its way out. The problem was that when I siliconed one of the seams between the slot drain form and the sink I was too aggressive with my finger pressure and had made it slightly concave. This had allowed the concrete to pinch in ever so slightly and bind the PVC slots in place. My hammering had finally squeezed the PVC through the slightly narrower gap. If I had used melamine instead of the PVC foam board it would not have compressed so easily, and worse yet it might have expanded due to all the excess water I threw on there to help the concrete cure. There is a good chance that I would have destroyed both the piece and sink form trying to separate them. Use PVC and make sure that your joints don’t bind the form in. W ELL ? The first thing I noticed about the piece was that there were no deal-breakers. There were quite a few pinholes, one small tear, and one partially eroded corner where sand had gathered, but nothing that I didn’t think I could fix. The next thing I noticed were the watermarks on the cream surface. There were pronounced white marks around the perimeter of both the vanity and the backsplash. Similar patterning surrounded other features such as the sink and toothbrush holder. I also noticed crisp lines where I pulled the tape after caulking the corners of the form with silicone. None of that was the look I was going for. F I NI SHI NG I was hoping to avoid a slurry coat or much grinding or polishing, but there was no avoiding it now. There were too many pinholes to ignore and the watermarks were just too pronounced. They didn’t really lend character, the just looked bad. I was hoping that I could just buff the watermarks off like Buddy Rhodes did with the twister pad in his video. I was unable to source twister pads locally, and didn’t have time to mail order so I went to my local Tool Depot to see what might work. I ended up purchasing Scotch-Brite pads of various textures for my angle grinder. I started with the finest pad on one of my small test pieces. The instant I turned the grinder on the pad flew off and hit me in the jewels. Hard. I thought I might not have put the pad on firmly enough so I tried again, this time with a face shield and the head of the grinder in a garbage can. Same result. Do not try this at home! Someone on the forum suggested that I buy a variable speed grinder at Harbor Freight and that it should last fine through my projects. I searched there and it seemed that they no longer sold them. I slinked back to Tool Depot where I picked up a used variable speed grinder. I learned later that HF does sell them but they call them “polishers.” The Scotch-Brite pads worked pretty well with the variable speed grinder. I was trying to polish the surface back to a uniform color while preserving the cream layer. I was fairly successful on the test pieces, but there was just too much contrast on the real piece and by the time I got the water marks off I was already pretty well through the cream layer. I ended up with something that looked like a very light grind.
The trouble was that the sand I had is fairly ugly with black and white grains showing on the surface. I had never noticed these in the sand itself but it appeared on the surface of the concrete. Strange. The other problem I had was removing the lines created where I pulled the tape masking the silicone. These detracted from the simple monolithic look of the concrete. I wanted them gone, but removing them meant removing the cream adjacent as well. The net result was a very uneven surface with a bit of a camouflage look. My wife was very nice and said that she liked it, but I was pretty unhappy with it. At this point I tried an acid wash in the hopes that would even it out. The wash removed the sheen but did very little to the coloration. Perhaps I did not leave the acid on long enough, it wasn’t concentrated enough, etc. It did, however, open up more pinholes. My wife said she preferred the glossier look so I agreed to buff it back to a gloss after refilling the pinholes. S LURRY I filled the pinholes with a mix similar to the face coat, but with just enough sand to stiffen it up. I also use finer #120 sand. It worked pretty well, but it tended to carve out when I polished since I didn’t give it enough time to cure before buffing. I also noted that it was a lighter, browner color than the piece. I thought this odd since many on the forum mentioned that the slurry was typically darker than the piece after polishing. S EALI NG I used the ICT system to seal. I won’t go into detail here since that may not be of general interest other than to say that it seemed to go fine for the most part. I will try to make a second post about that when I get the time. L ESSO NS L EARNED Next time I caulk a form I will try to avoid using masking tape on the silicone to avoid the lines they produce. If it is too hard to get good radius corners with an even bead and careful pull I might try using the masking tape, removing it immediately, then making one quick final pass to knock down the tape line. Do you pros have any hints? The radiuses of the silicone corners I made were too small. I will probably go a bit larger this time. I was at about 1/16” but 1/8 to 3/16 might work better to keep sand from getting trapped and to prevent pinholes. Avoid creating any concave areas in the form. My problem was the silicone between the slot drain and the sink mold. This made it extremely difficult to remove the form as described above. Make sure the face coat concrete is thin enough to spray all without thickening or clogging. I went on the thick side because I was afraid I would have trouble with the material hanging on the verticals if it were too thin. That may be the case, but at this point it sure seems like it is worth some risk to try it. The thick stuff was an epic fail. Finally, I will not add extra water next time during the cure. I suspect that this extra water might have snuck in between the form and concrete after the initial cure and caused a lot of the watermark trouble. We’ll see next time…
C O NCLUSI O NS This stuff is really hard to control precisely. I am all about process and consistency, but even after several experiments I was far from having predictable results much less being able to alter my technique to get any result I want. It you are not the kind of person that can embrace this initial lack of control, this may not be the medium for you. Travis Welken wrote an excellent post on his first experience and I found it to be spot on. The concrete wants a say in the result, and it is going to get it. Embrace that and you will probably be much happier for it. In the couple of weeks since I made the piece I have really grown to like it. It’s not what I tried to make, but that’s ok. Maybe it is better.
ADDENDUM When I looked for the post above I could not find it so maybe I forgot to post it. Anyhow, I have made two more pieces since the project above. Below is what I learned from them… T HI N M I X The thinner mix sprays much better and had no significant drawbacks. I used a milkshake-like consistency that poured easily from the bucket and this seems like the way to go. I am still getting clogging though and often have to clean the gun mid-coat. I have continued to use a thinner mix for the back coats as well. I mix it reasonably thick (pizza dough) for vertical faces, then add super plasticizer to thin it a bit more for the flat parts. At this point it is similar in consistency to wet-cast. I screed it level on the second coat and eventually do a light trowel to seal it. I am very happy with this process. SP I N U P F RO NT After about 6 batches I realized that my SP loads were remarkably consistent (within 5%) so I began adding ¾ of the SP with the liquids prior to adding any powders. This kept the mix wetted out and mixing went much faster. It also seems like this approach would help the materials mix more thoroughly and blend out any inconsistencies, although I didn’t notice any real difference in the results. Use this approach with caution so you don’t end up with soup! A I R P RESSU RE I also ended up increasing the air pressure. I bought a cheapo inline “regulator” from Harbor Freight which is just a leaky ball valve. It did some good stopping the flow when the trigger wasn’t depressed and allowing pressure adjustment on the fly. I ended up cranking the pressure up on the compressor and adjusting it to taste with this regulator. This would work great with a better regulator but I can’t find one that looks any better than the one I have. E DG ES The larger radiuses on the edges were also the way to go. Free-handing the caulk is also the way to go, but not for the faint of heart. You have to get a very consistent bead of caulk out of the gun and very consistent finger pressure or you could end up with a mess. The biggest problem was at the corners as I pulled my finger off. Any suggestions?
C URI NG I was right about extra cure water causing the extreme water marks on my first piece. I used only very fine misting on the backs of the next two pieces and the water marks were reduced by about 80%. T HE F I NI SH I am still not getting a nice, even, consistent color like I see in the work of others on this forum. Although better, I still have blotchy cream when I demold. I can even it out a bit with Scotch-Brite pads but I never get it very consistent. I really don’t see that ever happening with this process. It is not clear to me why not. It could be the locally sourced cement and sand or the spray process. Another problem that I had was discoloration and occasional crumbling at the edges of the form due to rapidly drying concrete. The very edge would flash off then want to delaminate and crumble in. I added an edge extender to the front edge in an attempt to solve this problem. The extender was simply an extra inch of melamine I attached flush with the actual form. After I had completed all coats and the concrete was just starting to set up I cut the extender away with a hack saw blade. I cut only toward the form in a shearing motion lifting the extender only enough for the blade to squeeze through. I only loosened the screws when they began to bind ahead of the blade. This process worked very well and left a beautiful edge. M Y S PRAY R I G This is the weak link in my system. When I spray it looks nothing like Buddy Rhodes GFRC video. His gun instantly lays down a nice, even, opaque layer of concrete from about a foot away. Mine rig just spits concrete at the surface eventually building up enough to cover it. I have to have my nozzle about four inches from the surface to get opaque coverage. I’m not sure if the problem is with the gun, the compressor, the aperture size, air pressure, mix thickness, or all of the above. I sure wish I could justify a nicer hopper gun but I am not absolutely sure that would fix the problem. I will keep experimenting and learn to make the best of what I have. Any suggestions in the mean time would be greatly appreciated!