|Paint booth built, rotisserie partially loaded|
|Time for "Practice Spray" with just water|
|Air Pressure Trigger Closed||61 psi|
|Air Pressure Trigger Open||45 psi|
|Paint Cup Weight Empty||165 grams|
|Paint Cup + Test Water (Start)||925 grams|
|Paint Cup + Test Water (End)||510 grams|
|Volume of Paint Cup to minor diameter||930 ml (31.5 fluid ounces, just under 1 quart)|
|Pyrex Cup Tare||280 grams|
|Pyrex Cup + 250 ml water||535 grams|
One 50 foot air hose, 3/8" ID, has a pressure drop of 16 psi at test flow conditions. Pressures measured with a pressure gauge attached to a brass tee. A male quick-disconnect was connected to the inlet of the tee. A 1/4" brass nipple and female quick-disconnect was attached to the outlet of the tee. This assembly was installed between air hose and quick-disconnect/swivel connected to spray gun. Pressures were measured without actually spraying any fluid. The pressure gauge assembly was removed from the line (hose connected directly to spray gun) for the actual spraying. It is possible that this results in a slightly higher pressure being fed to the gun, but estimated difference is small, probably less than 1 psi. End result is that the pressure to the gun would still be in the 40 to 50 psi range.
The idea of this exercise was to "practice spraying" and adjust the spray gun prior to actually spraying primer. This way, "mistakes" could intentionally be made without damaging the parts. That is, I could see how much spraying would be required (not much) to produce what would be the equivalent of a "light white haze" over the parts. I could then test the sensitivity of the spraying process by seeing how much longer I would have to spray to apply so much water that it would begin to run. Initial results indicate that a proper spray does not take too many passes, so I should make an active effort to hit the part from each direction with the minimum number of passes. This way the part will be covered without causing runs on a previously painted side while painting an unpainted side. Fortunately, it seems to take about 3 to 4 times that many passes to "over-do" it, so the sensitivity is low enough to be comfortable. Future experience with the actual primer should validate if this is a technique worthy of recommendation.
The first assumption was that the water would spray and adhere to the parts similarly to the Metal Prime. This assumption should be valid, since the solids should increase the viscosity and adhesion of the water if anything.
The second assumption was that the Metal Prime will spray and atomize like water. I have less confidence in this assumption, since if the viscosity and adhesion of the Metal Prime is higher than that for water, it will be more difficult to draw up the feed tube and atomize. Thus, the fluid feed adjustment may have to opened more. As a starting point, 1.5 turns open from fully closed on the fluid feed adjustment seemed to be a good setting for water.
The parts rotisserie was about half full of parts. This test spraying used 415 grams of water. This included spraying on the walls to adjust the flow, spraying all sides of the parts (rotating the rotisserie), and intentionally over-spraying the parts to cause runs. Weighing of water in Pyrex measuring cup of 250 ml of water gave 255 grams, which is within experimental accuracy to confirm a density of 1 gram/ml as expected. Hence, 415 ml of water were used, or 14 fluid ounces of water, or 0.877 pints. Roughly half the paint cup.
While I didn't measure precisely, drying time seemed to be on the order of 1/2 hour. We'll watch this more as we progress.
Metal Prime Density Check
Plastic Cup Tare = 15 grams
Plastic Cup + 8 oz Metal Prime = 315 grams
8 oz = 237 ml
Density = 300 grams/8 oz = 37.5 grams/oz
Density = 300 grams/237 ml = 1.27 grams/ml (27% higher than water)
Recommended cross-linker for Metal Prime:
2 grams (2 cc) per 16 fl oz of Metal Prime = 0.125 cc/fl oz = 0.00333 cc/gram = 0.00333 gram/gram (Change 1, March 2001: Crosslink to 4*0.00333 cc/gram = 0.01332 cc/gram)
As expected, the Metal Prime required a little bigger opening of the fluid control valve on the spray gun. It was increased 1/2 turn to be 2 full turns open from closed.
Will need to see if parts in middle of rotisserie are getting sufficient coverage or are just blowing and turning away from stream.
Transferred Metal Prime from can to paint cup by using cut down 16 oz plastic cup. Was able to rinse filter after filtering.
Had problems on first coat with rib stiffener angles moving in spray of gun, resulting in portions not being sprayed. Ran wire through angles to attempt to stabilize for second coat. Some improvement, twisted holding wires to hold angles better. Applied 4 coats to make up for portions missed on first coat.
4 coats shot without sanding in between. All of these parts are internal, so smoothness is not critical. Resulting finish was rough, like about 220 grit sandpaper. Surface can be easily smoothed with 320 grit sandpaper, but is not recommended, especially on small parts such as attach angles. Reason: Too easy to totally remove primer from corners/edges. Sanding would do more harm than good. Sanding should not be a problem for later exposed surfaces.
Later discussions with Doug Dodson and Gary Aldrich supported idea that primers are expected to have rough surfaces. Epoxy primer is an exception in that it flows out smooth, but then everything sticks to epoxy.
I told George Gennuso that I had no problems getting Metal Prime to spray like the problems he had described trying to spray Smooth Prime. George said that this matched what he had seen in discussions in the Pulsar builders group--that Metal Prime was easier to spray than Smooth Prime.
Practice spraying with water turned out to be a good simulation of the primer for technique development. Recommend this method be incorporated in the manual as a way to practice without fear of "messing up" for nervous first-time sprayers.
Rib attach angles were strung up as before, with two lateral wires passing through each angle. These two wires passed through the angles from opposite sides, with the result preventing the angles from rotating in the air stream of the spray gun. The vertical pieces and other pieces did not have a problem of moving in the air stream. This was tested by "spraying" the parts with no liquids in the gun, just air.
As a side note, Metal Prime doesn't stick to plastic Solo cups. After it dries, the resulting sheet just pulls off.
As before, the spray gun, minus the paint cup, was submerged in a bucket of water between coats. As is my custom, after the last coat was sprayed, dried, and inspected, the spray gun was cleaned by disassembly. Pipe cleaners (chenille stems--available at craft stores such as Michael's) are used to clean small passages and tubes. Since the gun is kept submerged in water, it is easy to clean the parts with just water. Poly Fiber Gun Cleaner is only needed occasionally, mostly for the top part of the inside of the paint cup where the primer can dry.
The table at the end of this document holds the data for all spray sessions. Some initial conclusions can be drawn at this time.
Using the system described above for adding crosslinker and thinning is easy to do and allows the use of arbitrary amounts of Metal Prime without a requirement to guess at amounts. This should result in good consistency of final product.
With comfortable temperatures (65°F to 75°F) and typical desert humidity (30% - 35%), each coat dries quickly, typically in less than 30 minutes.
While the paint booth will fog up rather quickly with overspray, waiting about 3-4 minutes will let the overspray settle or move to the exhaust filter, giving a clear atmosphere to look through again. This can be as short as the time required to rotate the rotisserie and reposition part dog tags.
The overspray will clog up the exhaust filter (a typical house heater filter) within one coat of spraying. The filter can be somewhat recovered by allowing this film to dry, then tapping or brushing with a wire brush to break the primer film. This has a down side of also removing part of the filter fibers, but it's better than replacing the filter each time.
Some of the overspray will settle on booth surfaces, drying to a form resembling snow. This could be seen as an interesting but useless advance--the ability to create snow without the effort of sanding!
Interestingly, the temperature inside the booth drops by 2 to 3°F during a spraying session. Probably just that "evaporative cooling" from all of the atomized water.
Current experience shows that one coat on a 3' x 15' set of parts takes about 500 to 800 grams of thinned Metal Prime. This has resulted in two sets of parts using 95% or more of a gallon of Metal Prime. These were all small parts, so a large amount of spray is overspray. Larger parts, such as skin panels, should have less "waste."
I am using a Tyvek coverall with integral feet and hood with goggles and a respirator for protection while in the booth. Vinyl gloves are used on the hands. Rubber bands around the ankles help to keep the feet under control. It seems to work well. Haven't had to clean up any additional clothing. I wear just a T-shirt, shorts, socks and shoes (okay, and underwear) under the coveralls, and it still gets warm in there, even when the booth temperature is 65°F.
Time spent suiting up and in the spray booth is running about 20 minutes. At least that much time is spent mixing the paint and doing other setup chores.
Ignoring the fact that I've never been too crazy about finishing, I like this product. I don't enjoy the spraying in a way that I look forward to it--it's more of a necessary evil. But compared to experiences using solvent based paints, such as enamels, this is so much nicer--even tolerable! I only have to use the respirator while in the booth spraying, and that's to avoid particulates. I don't have to wear it all of the time to avoid getting sick smelling the solvent fumes. Much, Much Better. The coverage is also satisfactory when following the directions. A few runs have been found, but I suspect these are from overspray while trying to coat some other area, and as such would probably occur regardless of the type of paint being used. These runs should sand out with no problem. Areas on large flat pieces cover very nicely. Just like Ray Stits inferred--follow the directions, and you'll get good results.
Sprayed primer on set 3 tonight. Metal Prime had just arrived that day, and looked like it had been shaken recently. Even so, I tried using a drill powered stirrer to make sure it was mixed up good. Had to use slow speeds on the drill so as to not "spray" primer all over the surroundings. Moved the impeller all around the bottom to dislodge any solids and up and down the sides. Biggest problem was close to the top, where spillage was more likely. Seemed to work okay.
I had been thinking about resolving a disconnect between the book and e-mail from Jon. The book says to use 3 coats, Jon had told me 2 for aluminum, 3 for steel.
I've been working on trying to find ways not to "waste" so much Metal Prime in overspray. Tried readjusting the spray gun. Found several points of interest:
The Poly(-)Fiber guarantee still holds! After doing this for a while, I started to forget the importance of using light, seemingly insufficient coats. I'm sad to say I started to listen to the voice of Bubba, and started trying to make that first coat cover in white. Sure enough, I started to get runs, especially where two flanges come together and concentrate the spray.
Later I would figure out that I should have wiped off the runs when I found them. At this time thinking I could just sand them out between coats, I tried to sand out the runs. Being only 30 to 45 minutes after spraying, the primer had attached itself to itself better than to the part. As a result, the primer would come off in sheets instead of sanding down like expected. Tests on parts that had been sprayed 3 weeks ago showed that the fully cured primer would sand down as expected. This behaviour makes me question arbitrarily sanding between coats. While okay for splatter blobs, it runs the risk of causing more problems than it solves.
Only two coats were shot on parts in general. Leftover Metal Prime after second coat was used to spray areas that needed a little more primer as a "third" coat. This should cut usage by 1/3.
Removed Set 3 parts and loaded up set 4. Areas that were thicker with primer (runs and such), especially around attaching wires, were still not fully cured and tended to pull off as described above.
Using more positions of the rotiserrie to try to be smarter about getting all sides of the parts.
Techniques for spraying ribs, to include proper coverage of flanges, turned out to be easy and effective. Must be careful not to over-do application in areas such as the web next to the flange, which picks up material from spraying the flanges. Challenge is to spray the rest of the web without overdoing these areas.
I am finding that the airborne overspray dries quickly, such that it settles on the surroundings already dried into a powder. The result looks something like someone went nuts with a talcum powder bottle. I actually see this as a benefit. Since it is impossible to contain all of the overspray, and one time forgetting to use the exhaust fan will result in overspray throughout the garage. After it has settled, cleaning items is nothing more than dusting them off. This was especially beneficial when I forgot to put the hood up on the paint suit. While the premature gray look was somewhat comical, it was removed simply by rubbing my hands through my hair. If I was using enamel, it would have stuck to everything, including my hair!
This was the first time to spray steel parts, specifically the wing strut attach straps. Rather than use the suggested Poly-Fiber C-2200 Metl-Sol, I used "Simple Orange Multi-Purpose Cleaner" (XCELL brand) which I purchased at Harbor Freight. The ingredients of this cleaner are d-Limonene and Sodium Laureth Sulfate. d-Limonene has been recommended as a biodegradable cleaner in The Leading Edge. No media blasting was required since no rust existed on the parts. After cleaning off the Sharpie pen marks with isopropyl alcohol (on the hot rolled surface), the parts were sprayed with d-Limonene and wiped down with a paper towel. It seems to have worked well, as dark "stuff" was found on the paper towel. After wiping down, the steel part was rinsed with water and then immediately blown dry with an air hose.
It quickly became apparent why an additional coat is recommended for steel compared to aluminum. The first coat sprayed on the steel did not seem to cover as well and similarly sprayed coats on alodined aluminum. So far, it seems that the primer is sticking well to the steel after treating with d-Limonene. I may try later with a scrap piece of 4130 tubing known to have that ubiquitous oil on it.
The Metal Prime interacts with Permanent Sharpie® Ink in an interesting way. All ink was "removed" with isopropyl alcohol from aluminum parts prior to Scotch Brite-ing and alodining. Ink was also removed from steel parts with alcohol before cleaning with d-Limonene. In most cases, this worked just fine. However, on a limited number of parts, apparently a small portion of ink remained even though it was not visible. The first coat of Metal Prime somehow brings out the small amount of remaining ink, making it visible through the primer. Sort of like invisible ink. I'm still determining how badly it bleeds through.
"Save it 'til the next coat..."
This new slogan seems to be working...
The temptation of Bubba and the dark side of "just a little more to give an even white coverage" while spraying Metal Prime is incredibly strong! Luke Skywalker thought he had it bad...
After quite a few sets of spraying with runs here and there, I think I may be finally developing the techniques that may avoid (most) runs.
1. Spray lightly--quit when the coat only looks half done. Sure, I know you said this in the book, but it takes incredible self control to actually do (especially since I don't enjoy all the work of spraying (hard work to do correctly) and the idea of "another coat" can be depressing...). As I get more used to this system, it shouldn't be so difficult. It's what we military flying folks call "negative transfer of learning"--what worked in your last airplane doesn't work in this one...
2. Allow the parts to dry in a horizontal position--typically the position the part would be in if set on a table. With my limited space in the paint booth and lack of gorilla arms, it is usually easier to shoot the parts in basically a "vertical" position, as though painting a wall. After I'm done shooting, I rotate the rotisserie into the "horizontal" position. That way the paint just sits there instead of running down a part with gravity, picking up more paint as it goes along, eventually creating a big, ugly run.
3. I'm getting better with experience at deciding what spray pattern to use and adjusting the flow of paint to keep it under control. Too fast a flow and it becomes uncontrollable. Also, the wider the fan pattern, the greater flow required to get the same percieved "rate" of paint application.
I'm sure none of this is new to you and your paint gurus, but I find it useful at work to listen to the students as they go through the learning process and discover things I already know. It reminds me of what may seem obvious now, but at one time was a big mystery. The teacher must know the path the student will take in learning to be able to guide him down that path.
Painting efficiency seems to be increasing--it looks now like I may be able to just finish up the parts I need to do with the remaining Metal Prime left. If not, you'll hear from me again. Actually, you'll hear from me again anyway...
I've mentioned it before, but now I'm convinced it is happening.
Various pieces of aluminum, particularly bar stock, come with printing on them to identify the stock. Because the surface is not polished like alclad sheets, it is more difficult to get the writing off with alcohol from the bar stock than from the sheets. In addition, there are the various identifying marks applied with a Sharpie pen during the construction process.
Prior to alodining, I removed as much of these marks as possible with alcohol. Any remaining marks were then "removed" while Scotch-Briting the surface to clean off the aluminum oxide for alodining. Prior to going into the alodine tanks, the markings appear for all practical purposes to be gone. After alodining, I still could see no trace of any marks.
However, after spraying with Metal Prime, many of the marks, mostly the ones that would not come off with just alcohol, magically reappear! This tells me that some of the ink was still there, and must somehow react with the Metal Prime. These marks are still visible after the third coat of Metal Prime. Remember, we have exorcised the spirit of Bubba and are putting these coats on lightly, but it is still a white coverage by the third coat. Additionally, the marks are as visible after the last coat as after the first coat. The same thing happened with marks on hot-rolled steel (again, not a smooth surface).
Anyway, two questions come to mind:
1. Can you or your boffins tell me what's going on here? What is causing the "ink" to reappear?
2. Is Top Gloss or whatever I'm going to eventually cover the primer with formulated differently such that it will actually cover this "ink"?
I annoy Tech Support technicians--I never call with the easy questions--only the tough ones.
This weekend I primed the first set of 4130 welded steel parts.
I started by sandblasting the parts. I used my Sears sandblaster with 90 psi air pressure. For sand I used 60 grit silica sand. I did the blasting outside "open loop" with no attempt to recover the sand. I was impressed by how easily this removed the scale left from the welding. The blasting also removed the black oil surface that normally protects the tubing, leaving everything with a nice, uniform gray surface. I used a respirator and goggles, which were an absolute requirement. A noticeable amount of the sand bounced back in my face regardless of what direction I held the sandblaster. I had no trouble with sand in my eyes or breathing any sand. My hair, socks, and everything else was another story--like a day at the beach.
I further cleaned each part with d-limonene cleaner, followed by a water rinse and compressed air drying. The parts were then strung up in the paint booth on the rotisserie.
I used the same procedure to spray the parts that I had used on the aluminum parts. Ever mindful of the spirit of Bubba, I was conscious of not spraying it on too thick. However, I was surprised that the spray not only seemed to stick better to the steel parts (perhaps a rougher surface to cling to?), but also seemed to cover it quicker. That is, the same spray that gave aluminum the look of a dusting of white looked almost like a finish coat (completely white) on steel. At first I was worried that I had overdone it and would have runs galore, but these did not happen.
It seems to work very well on steel. As always, one of the best parts was the ease of cleanup.
|Description of Operation||Date||Paint Cup and Leftover Primer Tare||Paint Cup + Added Metal Prime||Net Metal Prime Added||Crosslinker Amount||Volume Metal Prime||Water to Add||Total Weight Paint Cup, Metal Prime, Water||Start Temperature||Start Humidity||Time||Final Temperature||Final Humidity||Time||Weight Final Paint Cup, Metal Prime||Net Thinned Metal Prime Used||Time to Spray|
|1st Coat, 1st set||24-Oct-98||165||995||830||2.8||654||131||1126||71||31||15:27||68||31||15:44||605||521||0:17|
|2nd Coat, 1st set||24-Oct-98||610||820||210||0.7||165||33||853||70||34||17:22||67||35||17:39||295||558||0:17|
|3rd Coat, 1st set||24-Oct-98||295||870||575||1.9||453||91||961||70||34||19:33||67||34||19:45||300||661||0:12|
|4th Coat, 1st set||24-Oct-98||305||820||515||1.7||406||81||901||67||35||21:24||63||36||21:42||225||676||0:18|
|1st Coat, 2nd set||31-Oct-98||165||880||715||2.4||563||113||993||67||30||14:52||65||30||15:16||175||818||0:24|
|2nd Coat, 2nd set||31-Oct-98||175||1025||850||2.8||669||134||1159||68||30||16:25||65||30||16:44||365||794||0:19|
|3rd Coat, 2nd set||31-Oct-98||365||940||575||1.9||453||91||1031||66||30||17:54||63||30||18:11||485||546||0:17|
Revised -- 19 March 2001