How To Stop Profile Warpage
Faster! Every company wants to run faster to get more product out on the same production line and from the same amount of labor. Plastic profile extrusion companies are no exception. It is easy to speed up the extruder to push more pounds or even to buy a larger extruder to get more output. However, when extruding plastic profiles, the output is usually controlled by the cooling of the profile and the ability to hold the part in the correct shape while it is being cooled. It is hard enough to cool simple shapes like round pipe and tubing faster but the difficulty increases when the complexity of the profile increases. Window profiles and other complex parts are very difficult to cool uniformly, and if the parts do not cool uniformly warpage and bow is the result.
Like most materials, plastics shrink as the temperature of the plastic decreases, but they usually shrink a lot more than other materials. Plastics shrink at one rate when they are in the solid (frozen) state, but they shrink much more when they are still soft or in the molten state. The problem for the profile extruder is controlling this shrinkage when cooling the hot plastic, coming out of the extruder, all the way down to room temperature. Let’s take the simplest example of a flat sheet where one side cools faster than the other. When still soft both sides are shrinking at the same rate. Even if one side is cooling faster and shrinking faster the other side is still pliable enough to come along with the other shrinking side. However, once one side cools past the crystalline temperature or its glass transition temperature, two things happen. First, that material stiffens and is no longer pliable enough to follow the other side and the rate of shrinkage goes down significantly. It is as if the stiffened side is no longer shrinking while the other pliable side continues to shrink. Therefore, as the pliable side continues to shrink it is pulling on the stiffened side and causing a bow in the direction of the side that cooled last. In this example, and in other simple profiles, the part will bow in the direction of the material that cooled last. In more complex profiles the parts may twist, distort, or warp in all types of fashions depending on which sections of the part cooled last. We’ll cover more on this later.
In addition to this problem is the fact that plastics are good thermal insulators, meaning that they don’t transfer heat very fast. That means it is difficult to pull all of the heat out of the part in the first place, let alone doing it uniformly. Thermal conductivity is a measure of how fast materials transfer heat. Steel has a thermal conductivity of 43 while Aluminum’s higher heat transfer is 250 and most plastics are way down at values between 0.1 and 0.3.
TYPES OF COOLING AIR COOLING
Considering these problems with cooling profiles it should not be surprising that historically profile extruders often used air to cool parts.
Air racks are simple tables or frames with plates / guides and fixtures that hold the part in shape as it is being pulled slowly across the table. Fans are generally used to improve overall cooling while compressed air jets are added where specific additional cooling is required. Metal fingers, wires, and jigs attached to the table with clamps or vise grips are used to push the part into shape as it cools very slowly.
Air is very inefficient, meaning SLOW, which in this case is good because slow gives the operator time to make adjustments and get the part just right without warping or other distortion. Complex profiles or parts with different wall thicknesses on different sections of the part may need customized cooling. The operator can direct more cooling to where he needs it with compressed air nozzles or retard cooling in other areas by insulating a section to keep it from cooling too fast. Since thicker sections cool more slowly than thin sections, specific actions must be employed to avoid warp. The operator will need to direct significantly more cooling on thicker sections to get them to cool to the same temperature at the same time as thinner sections on the same profile. Likewise, inside a U-channel or simply an inside corner will cool slower than an outside corner and will require more directed cooling. Output rates are limited to between 100 – 250 lb./hr. using air because it is so slow.
Even today, some may still use air cooling when:
- Profiles are very complex
- Using materials with very different thermal conductivities
- Size of production runs do not justify more expensive tooling
SUBMERSION WATER COOLING
When higher output rates are required, then cooling with water is used. There are many ways to run a part through water depending on many variables.
For very simple shapes the part can be extruded over the top of a long water tank and be pushed down under the water by rollers or sizing plates. This can only be used for parts where it doesn’t matter that the bottom of the part hits the water first (and is cooled first) while the top comes down into the water an instant later.
Extruding larger or more complex shapes straight into the water tank is a great idea that runs into the simple problem of gravity pulling water out of the tank through the hole that the part needs to go through into the tank. Even small gaps between the sides of the part and the sides of the entrance plate will allow water to leak out. This problem is usually solved by applying vacuum to the entire inside of the tank to hold the water in. Of course, this requires a special tank that is strong enough not to collapse from the differential force of vacuum on the inside and air pressure on the outside of the tank.
Another option is to make a small vacuum sleeve around the entrance to suck off any water trying to flow through the gap between part and entrance plate. More recently, profile extruders will place a dry vacuum calibrator in front of the water tank to accomplish the same thing. This vacuum calibrator can be as short as 3” for less critical profiles or as long as 10 feet for parts that have to be hardened to very precise dimensions before going into the water tank for more cooling. Dry vacuum calibration is not as efficient as water cooling but it is the price that must be paid when tighter control of the dimensions is required.
Water Temperature Choices
It’s pretty obvious that vacuum tanks are totally closed. Even with an open water tank it is very difficult, if not impossible, to get into the tank to place fingers and jigs to push the part into shape as is done on an air rack. It is also difficult to direct cooling water or to insulate sections of the part from cooling. However, it is possible to reduce the efficiency of cooling (i.e. slow it down) to mimic the more uniform cooling possible with an air rack by heating the water. This is often done with parts that have a strong tendency to warp and especially with higher temperature engineering materials. In this case a temperature control unit is required to control the temperature of the water at a set value. The higher the water temperature is the slower the cooling and therefore the easier it is to achieve uniform cooling. Controlled temperature water between 80° F and 130° F is often used in the initial tank until colder water can be used to complete the cooling. Of course, with the desire for speed, the colder the water the faster the cooling, so most profile extruders will use chilled water at temperatures between 50° F and 55° F whenever they can.
Water Flow Characteristics
Even though immersing the entire profile in water gives faster and more efficient cooling it may not be the best cooling method. Unless the water is being agitated to give turbulent flow around the part, then the layer of water next to the part will heat up and that hot water next to the part will slow down the cooling. The same phenomena may occur on simple shapes like round pipes or tubing to cause uneven cooling and bowing. We all know that heat rises and heated water is no exception. This is great for the water next to the vertical surfaces of a part going through the water. The water is heated by the part and this heated water will rise along the part drawing cold water behind it to further cool the part with a continuous renewing of cold water against the part. However, heated water on the bottom surface cannot rise as easily because the part is in the way. It does slowly move up and draw cold water behind it but less efficiently than what is happening on the sides. The top is more of a problem because even though the heated water is not obstructed from moving up and away from the part, the only water that is drawn in to replace it is the heated water moving up the sides of the part. The top is not cooled as fast and pipes or other parts will generally bow up (bend upwards). Sizing plates in the tank help break up this flow but only allow cold water onto the top of the part immediately after the sizing plate. Turbulent circulation of water in the tank significantly helps with this problem.
Spray cooling is an improvement over immersion cooling and another way to answer the cooling challenge. Spray nozzles are evenly distributed around the part and down the tank to ensure a constant replenishment of temperature controlled water to the surface of the part. This spray also ensures more uniform cooling by spraying water equally into U-channels and inside corners compared to outside corners and straight walls. Parts with a simple cross section can be sprayed with chilled water and run at high rates of production. The challenge of uneven wall thicknesses still needs to be addressed separately. If spraying cold water alone is not sufficient to achieve the uniform cooling that is needed to avoid warping, the water can be temperature controlled to slow down the cooling and reduce or eliminate warping. Water is required in a sufficient volume to create the turbulent flow in the tank that is needed to break up the insulating layer of warm water.
Some people claim that spray cooling is significantly better than immersion cooling due to the evaporative cooling effect. This is where the water sprayed onto the hot part is quickly turned to steam and evaporates carrying off significantly more heat than the water can carry off when immersed. While this effect is real, it is only true when the surface of the plastic is above about 250° F. This only happens in the very first seconds or even tenths of seconds of the part entering the cooling tank. With the high efficiency of cooling of the water and more importantly the very low conduction of heat from the plastic to the surface, the surface temperature quickly drops below 250° F. and stays there so that no more evaporative cooling occurs. Still, the constant replenishment of cold water to the surface is an improvement in the efficiency of the water cooling, with the added benefit of not requiring vacuum to hold the water inside the tank. Spray cooling does offer more uniform distribution of cooling water over the surface as well as continuous replenishment of cold water on the surface with the added benefit of using much lower flow rates of water.
So, the plastic part will tell you when it is not being cooled uniformly by bowing, warping, or distorting. With simple shapes the part will bow in the direction of the wall or section that cooled last. In more complex shapes the contortions may not be as easy to figure out with as many as 6 to 10 different wall sections cooling at different rates. Directing more cooling to sections that obviously would cool slower because they are: thicker, inside corners, otherwise shielded from circulating or spray water will lead to control of warpage. Now the trick is to speed it up and solve the problem all over again.