Assesment of Paper Machine Noncode Dryers for Upgrade

Recently, we've had several assignments from pulp mill clients who have encountered concerns from their insurance companies about the continued operation of their old, noncoded cast-iron paper machine dryers, some of which date back to the early 1900's. In most of these cases, the mills have been operating the dryers for many years without incident, but it seems there's been an increased awareness among the insurance companies about the potential hazards of continuing to operate these older pressure vessels, and as a result, the insurance companies have begun to require that the mills drastically reduce the operating pressure.

So what? Well, for our nonpulp-mill clients, let's start with a brief description of what a dryer is and why it's important to the mill. Basically a dryer is just what its name implies: a device for drying the moisture out of paper as it's being processed from its wet pulp stage to its final, ready-for-market paper stage.

The dryers are cylindrical, rotating, cast-iron pressure vessels heated with saturated steam, typically at a pressure of 30 to 60 psi. There's usually somewhere between 30 and 80 dryers in a single paper machine, and as the wet paper sheet successively passes across the heated surface of each dryer, it loses moisture along the way.

The amount of paper a paper machine produces is governed by its drying capacity, which in turn is governed, among other things, by the number of dryers and how hot they are. And the temperature of the dryers? That's determined solely by their operating steam pressure, which is where the insurance company enters the picture for the old, noncode dryers.

Before proceeding, let's consider just what we mean by a "code" or "noncode" dryer. Dryer drums, when operated at pressures over 15 psi, are classified as pressure vessels, and since the 1940's, they have been subject to the provisions of the ASME Boiler and Pressure Vessel Code. The pressure vessel code governs all aspects of the design, materials, and fabrication of pressure vessels to assure they comply with good engineering practice and appropriate factors of safety. When a dryer is built in a code shop, it gets a nameplate and appropriate documentation certifying that it complies with the requirements of the pressure vessel code, and this documentation is then used to validate the operating pressure of the dryer.

The problem with the noncode dryers is they were built before the detailed requirements of the code came into existence, and in many instances, little or no documentation exists about the materials or quality control techniques used to cast the dryers; without this information, it becomes difficult to calculate how high a pressure the dryers can safely be operated at. So what's an insurance company to do? Since their goal is to protect life, limb, and property, they take a conservative approach with an assumption that all the noncode dryers are made with a low-strength cast iron, and consequently, they calculate a correspondingly low maximum allowable operating pressure.

For the purposes of the calculation, the tensile strength of the cast iron is assumed to be only 13,000 psi; this results in a low operating steam pressure, generally less than the dryers have been operating at in the past. As we already said, when the pressure is reduced, the temperature goes down, which of course reduces the drying capacity of each dryer. To compensate, the rotational speed of the dryers must be reduced so the paper spends more time traversing them in order to accomplish the same amount of drying.

What this does, of course, is reduce the through-put of the paper machine, which naturally has a negative impact on the company's "bottom line." So what can be done? That's where we come in.

In most instances, determination of cast-iron tensile strength from a sample removed from the dryer shell is not practical unless the mill has a spare dryer to sacrifice. And unlike carbon steel and certain other metals, cast iron, because of microstructural variations, does not exhibit a consistent correlation between hardness and tensile strength; thus, a simple hardness test is not sufficient to estimate the strength of the material. Instead, the strength of the material depends in a complex manner on both its hardness and microstructural composition.

Over the years, there have been a number of methods proposed to estimate the tensile strength of cast iron on the basis of properties that can easily be measured in the field. Almost all of these methods involve hardness tests; a somewhat less common method utilizes the acoustic velocity in combination with hardness. Unfortunately, these methods all reveal a substantial scatter in the data, to the extent that in some cases the estimated tensile strength can be off by a factor of two or more—sometimes in the wrong direction from a safety standpoint (i.e., the actual strength is less than the estimated strength).

We have found a more reliable procedure is to utilize a metallurgical evaluation of the material in combination with hardness tests. This is a substantially more difficult undertaking because a metallurgical evaluation requires on the order of two hours or so per spot examined, compared to only a minute or two for a hardness test. (Of course, on the other hand, it's a whole lot less difficult than replacing the dryer with a new one.)

The metallurgical evaluation involves preparing a spot by sanding and polishing it through successively finer stages until a mirror-like surface finish is achieved, after which the surface is etched with a dilute acid solution to reveal the metallurgical grain structure. The spot can then be examined with a portable metallurgical microscope, usually at a magnification of between 100 and 500X.

One time-saving procedure we usually use is to replicate the grain structure with acetate tape, rather than examine it directly with the portable microscope. That way, we don't have to take the time to set up and mount the microscope on the dryer, which can be a time-consuming and difficult task in the field. By bringing the replicas back to the laboratory and examining them in a clean, quiet, vibration-free environment, we obtain better results anyway, plus, we don't spend the client's precious down time doing the examination in the field, but instead, we can replicate additional spots for assessment.

Sometimes the client has spare dryers available. Ideally, these are dryers no longer suitable for service, so they can be subjected to various destructive tests. In those cases we're able to go a step further than just estimating the strength of the material based on its microstructure and hardness, and we're able to actually measure the strength via laboratory tensile tests.

Several years ago one of our Canadian clients was facing this pressure-rating problem for his noncode dryers. The client owned an out-of-service spare dryer that he was willing to sacrifice in the interests of science, so he shipped it to us, whereupon we proceeded to conduct a full-size hydrostatic test to destruction at an abandoned rock quarry. Well, that is, it was supposed to be a test to destruction, but unfortunately, the heads were corroded to the extent they wouldn't seal properly (which is precisely why the dryer was no longer in service); thus, we were only able to achieve a pressure of about 250 psi before the leak rate exceeded the capacity of our pressure pump, and we were unable to blow it up. (Yes, the latent "little kids" in all of us grown-up engineers were quite disappointed when we didn't get the opportunity to see the pieces fly!)

Anyway, we then conducted a metallurgical evaluation of the material, along with tensile tests on coupons that we removed from the shell, and from this we were able to provide the client with a solid engineering basis for continued operation of his other dryers at higher pressure than would have otherwise been allowed by the insurance company.

Of course, these types of engineering assessments need to be done in conjunction with thorough nondestructive inspections of the dryers to determine their present condition. These inspections are done in accordance with established TAPPI (Technical Association of the Pulp and Paper Industry) guidelines and entail ultrasonic thickness testing of the shell (and sometimes the heads), magnetic particle inspections of the shell and heads for cracks, and ultrasonic inspections of the head bolts.

In a few of our recent noncode dryer assessment projects, we were pleasantly surprised to see just how good the metallurgical characteristics were on the old cast iron dryers. As a result, the mills were able to utilize nearly a hundred percent increase in the estimated strength over the assumed strength, giving these old workhorses a new lease on life with improved productivity.

Studies in Decrepitude

In our spring newsletter last year we described our investigation of a Chinese statue to determine if it was of recent or ancient origin. Because the material itself doesn't degrade over time (other than corroding from the surface), there aren't any handy aging tests that could be run, such as Carbon 14 dating, which is used extensively on biological samples. Instead, we relied primarily on the details of the statue's fabrication and material composition and our knowledge of ancient metallurgy practices to establish its age.

We also work at the other end of the investigative spectrum by considering the effects of the environment over time on newly manufactured items to predict how long they will last. The multiple assaults of heat, freezing, ultraviolet light, humidity, rain, and environmental contaminants like salt and acids (such as sulfur dioxide) all take a toll on nearly everything we touch. Frequently, our jobs involve sorting out and measuring the effects of these agents to predict the longevity or durability of new products.

Typically, our first task is to find out which of the multiple agents attacking the product is likely to do the most damage. That is, we first find the weakest link. If that can be done, then the testing that follows may be fairly straightforward and able to predict longevity, though sometimes, you need to watch out for synergisms between the different corrosive or weathering agents.

A recent case in point was a request to simulate one or two years of outdoor weathering on wood. A variety of methods are available, such as ASTM D 1037 "Standard Methods of Evaluating the Properties of Wood-Base Fiber and Particle Board Panels." This test involves soaking, steaming, freezing, and drying a test object multiple times and then evaluating it for deterioration. Of course, in typical ASTM fashion, no guidance is given on how these treatments might relate to real-world aging. For that, you need to read the latest research articles in forest product journals.

There you find a multitude of studies, some of which span thirty years or more, correlating accelerated aging test methods with the results of actual outdoor exposure. Unfortunately, none of the accelerated aging tests do a particularly good job of uniformly simulating the results of real exposure. Instead, the results show that the degree of degradation is often quite a bit different for different properties.

For example, one accelerated aging test is boil-dry cycling. Just as the name implies, this test involves one or more cycles of boiling and drying of a test specimen. For one brand of oriented strand board, boil-dry cycling resulted in swelling of the board thickness equivalent to over 5 years of outdoor exposure. However, the change in the modulus of rupture was only about the same as for one year of outdoor exposure. With another test method, cycling resulted in a change in the modulus of rupture equivalent to only one year of outdoor exposure, but the change in the internal bond strength was equivalent to about 5 years of outdoor exposure.

So as you see, even with the multi-year studies available in the literature, it's difficult to find a single accelerated test that will accurately simulate the effects of real weathering uniformly on all properties. Now imagine you have a new coating material to test, one which doesn't yet have any test results to go by. Without the thirty years of government-sponsored research studies to guide you, you won't know which accelerated aging test to choose, how long to run it, or how the various material properties are affected individually.

Frequently our assignments involve evaluating a material that has already degraded prematurely, rather than trying to predict in advance whether it will degrade prematurely. Usually we try to focus on simply finding and testing the weakest link. For example, we were recently asked to determine the cause for premature deterioration of some underground electrical cables which had shorted out, causing a serious fire. Samples submitted from an area away from the fire had brittle, cracked, and pitted surfaces, obvious signs of premature deterioration (see photo left). We immediately suspected overheating from electrical overloading.

But before proceeding any further, we first needed to know what the materials were. Our analyses showed the cables had a thick inner coating of polyvinyl chloride elastomer mixture with added dialkyl phthalate plasticizer and a thinner outer coating of nylon. At the interface, the two layers were stuck together in an abnormal fashion.

Knowing the materials and circumstances of use, we looked for a critical change only heat would produce. Since heat would drive the plasticizers out ward, one would expect excess plasticizer to collect at the nylon interface, resulting in degradation and solvent gluing at the interface. By subjecting portions of a nondegraded sample of the cable to various time-temperature profiles, we were able to reproduce the same type of sticking together found in the subject samples. Further, we were able to identify the temperatures at which this occurred, and these temperatures were consistent with electrical overloading of the cables. Thus, our testing indicated electrical overloading was the cause of the deterioration, rather than say, some type of chemically induced environmental attack.

MEI-C People

Chris Gerdes, Engineering Technician, and his wife, Dana, had an 8 lb, 14.5 oz baby boy on 11 February 2002. Cedric Evan Gerdes was born 22 inches long at 5:55 am and is doing fine. Congratulations, Chris and Dana!