Sponsored By

March 1, 2001

7 Min Read
The Materials Analyst, Part 41: Contamination derails the welding process

0301i54a.jpg

This series of articles is designed to help molders understand how a few analytical tools can help diagnose a part failure problem. Michael Sepe is our analyst and author. He is the technical director at Dickten & Masch Mfg., a molder of thermoset and thermoplastic materials in Nashotah, WI. Mike has provided analytical services to material suppliers, molders, and end users for 15-plus years.

In spite of the fact that polymer blends have become an important part of the plastics industry, most polymers do not naturally mix into useful compounds. There are exceptions, of course. The compatibility of PPO with high-impact polystyrene was discovered in the 1960s and turned a barely processible material into a burgeoning product line that we know today as Noryl. With care, ABS can be blended with SAN to vary the balance of stiffness and toughness in a compound, and some types of nylon can be mixed to improve surface finish and ease processing. All of these mixtures can be melt blended, which means that simply dry tumbling the compounds together and remelting the mixture in the injection molding machine will produce a usable part with a relatively homogeneous structure. 

But most commercial blends rely on supplemental chemistry that cannot be duplicated on the production floor. Polycarbonate/polyester alloys and nylon/PPO alloys are well known and are used effectively every day, but if a processor attempted to produce such a blend by simply mixing and molding the primary polymers, he would obtain a very different and disappointing result. The materials may delaminate, produce changes in gloss and other appearance characteristics, and exhibit reduced properties. When this happens we don't call it blending, we call it contamination. 

Welding Failure 
In robust part designs, low-level contamination may go unnoticed during the normal inspection and testing that accompanies the molding process. But if the molded part is involved in secondary operations, the problem may become evident later after value has already been added to the product. This was the problem faced by our client in this month's story. This manufacturer has a wide product line based on parts made from 20 percent glass-fiber-reinforced PPO/HIPS, known commercially by the brand name Noryl. 

One of the key advantages of Noryl is its ability to be effectively hot plate welded. Properly done, hot plate welding of Noryl can produce a bond that is actually stronger than the parent material. In the case of this client, a bond of good integrity is critical to the function of a consumer product that must operate trouble-free for decades. 

In robust parts, low-level contamination may go unnoticed during normal inspection, but become evident during value-adding processes.

But on this particular day, parts were simply refusing to weld. When they did stick together, they were easily separated at much lower forces than normal and the interface showed no signs of adhesion. The material had obviously been heated and had deformed, but no strength was created. The client had been through the heat welding process and had found no abnormalities with the equipment, so it supplied us with a good and a bad assembly consisting of a side plate and a valve body. 

The material at the interface of the bad assembly looked like it was low in glass fiber, so we first ran a glass content determination on both sides of the good and bad weld using thermogravimetric analysis (TGA). Figure 1 shows one of the results that was typical and gave the lowest glass content, 18.2 percent. While this is slightly below the nominal of 20 percent, it is within the specification range for the raw material—no smoking gun there. 

0301i54b.gif

Figure 1. TGA showing glass content of bad weld area.

With the glass content in the normal range it told us that if there was contamination it must either be a low percentage, or it involved another material with a comparable glass content. Our client had reported no unusual deviations in part weight, which further supported the idea that any foreign material would have to be present at a low percentage. 

Testing with DSC 
This is the point where the analyst has to make a decision about the best way to find a potential problem with material composition. The two major tools available for looking at polymer composition are differential scanning calorimetry (DSC) and infrared spectroscopy (IR). In detecting and analyzing a mixture of materials, the effectiveness of a technique is determined by its ability to highlight differences in properties. Since Noryl is an amorphous material, it provides a very simple result when tested for transition temperatures by DSC. This means that almost any contaminant with a melting point will be measurable by this method. For this reason, we chose DSC in our first attempt at diagnosing the problem. 

0301i56a.gif

Figure 2. DSC comparing material from a good and bad weld area.

Figure 2 shows a comparison of the side plate from the good and the bad assemblies. These samples were taken directly from the weld area. The good part shows only the step change associated with the glass transition of the PPO/HIPS blend. However, the bad part shows a crystalline melting point at 220C, a temperature consistent with both nylon 6 and PBT polyester. 

A DSC comparison of the valve bodies in the two assemblies revealed that both of these parts were free of any abnormalities. While this result points to contamination only in the side plate, the fact that the samples had been taken from the weld joint itself left a lingering question about the source of the problem. It was conceivable that the contamination could have transferred from one part to the other during the heat welding process. 

0301i58a.gif

Figure 3. DSC comparison of a good and bad sample away from weld area.

To clarify the situation, samples were taken from areas of the molded parts far removed from the weld zone. Figure 3 shows the DSC result for both the side plate and the valve body involved in the defective weld and confirms that only the side plate was contaminated. The intensity of the melting event associated with the contamination was approximately the same away from the weld and at the weld, suggesting that the contamination was homogeneous throughout the part and was therefore probably the result of mixing before molding. We estimated the level of contamination at 5 percent. 

We mentioned that the melting point is consistent with either nylon 6 or PBT polyester. If we needed to know the specific type of polymer that caused the contamination, infrared spectroscopy would be the tool of choice to make this determination. In this case, however, the client had enough information just knowing that the material was contaminated. 

Secondary Step Revelation 
In past cases involving contamination, the search for a contaminant began because the part in question was obviously compromised either cosmetically or in terms of physical properties. In this case the part design was robust and the contamination level was low enough to hide any problems with either performance or appearance. Without the secondary operation, the problem may have gone undetected and the parts would have gone into the field. Whether or not it would have caused a later problem is hard to say. The application involves constant contact with water, and the PPO/HIPS is selected for its low water absorption and excellent dimensional stability. 

If the contaminant were nylon, this clearly would have changed the moisture absorption characteristics of the part and may have compromised the dimensional stability, perhaps leading to other problems in the field. In this case, the secondary operation served as a built-in quality control check on the material, and relatively simple analytical tools isolated the problem. 

Contact information
Dickten & Masch Mfg. Co.
Nashotah, WI
Mike Sepe
Phone: (262) 369-5555, ext. 572
Fax: (262) 367-2331
Web: www.dmanalytical.com
E-mail: [email protected]

Sign up for the PlasticsToday NewsFeed newsletter.

You May Also Like