Testing Materials for Use in
William T. (Chip) Reynolds
Historic Ship Maintenance and Renovation
Captain, Half Moon
1507 Amherst Road
Hyattsville, MD 20783
Those responsible for maintaining and operating historic vessels, either original or replicas, are constantly torn between two poles. On one hand, we constantly seek to develop innovative new methods of ship maintenance that are environmentally and occupationally friendly, inexpensive, durable, and authentic appearing. On the other hand, as conservative traditionalists, we are loathe to change from methods that have been proven over the years to work well, and actually are authentic.
The purpose of this paper, being presented at the request of one of the conference organizers, is not to debate the merits of modern materials versus historic materials. Properly used, both modern materials and traditional materials have specific applications for which they are suited. In some cases, one may be substituted for the other; in other cases, no substitutions are appropriate. The focus of this paper is how one develops controlled tests of materials to determine their usefulness (or lack thereof).
Testing of Materials
Those responsible for maintaining historic ships are all in the business of testing materials. Unfortunately, most of this testing is anecdotal, non-formal and lacks proper controls. We tend to use a single product and observe its behavior, and then draw conclusions. Very few of us have line items in our budgets for research and development. In most cases, we are lucky just to have line items for salaries and basic materials. Consequently, we need to find economical ways to incorporate formalized testing into our normal course of business.
We also tend to operate on too short a time frame, attempting to resolve problems immediately at hand without planning for the long term and incorporating materials testing into our plans. For example, I am presently scraping the spars of the Half Moon (a replica of the Dutch Jaght that Henry Hudson sailed to North America in 1609) and need to determine what type of finish I will use on these spars. I must re-rig the ship by mid May, and consequently will not have time to test different products and different formulations. So I must make a decision based on experience and recommendations from trusted colleagues rather than empirical testing. However, in the process of refinishing the spars I will apply tests of different finishes that can inform decisions in the future when the ship is on a normalized schedule of routine maintenance.
Objective testing of materials is well within the means of any ship operator who has the capacity to plan ahead, regardless of their budget and financial situation. The key is not so much having lots of money to spend, but getting 'ahead of the curve' in your planning so as to anticipate the types of materials you may want to test, and fitting that into your schedule. In the case I noted above, no matter how much money I spend, I simply cannot test finishes for the spars that I must coat in the next few weeks. But in finishing these spars I can, in a controlled and documented fashion, use several different preservatives and finishes to determine how well they perform. In a year or two, that data will be very useful, and will cost very little to obtain.
Several elements are needed for an objective test:
- Develop a clear concept of the matters you want to test.
- Limit the variables to be tested.
- Establish a control.
- Define and adhere to clear procedures.
- Ensure that all your samples are treated the same.
- Document your procedures, observations and results.
- Quantify, measure, weigh, time, photograph, record.
- Write your findings in a permanent and retrievable location, and if possible publish them in a publication that is accessible to others.
1. Develop a clear concept of the matters you want to test: You need a clear statement of what it is that you are going to test. It is important to write this central concept in no more than a sentence or two. "I want to determine what is the best method of treating my iron rings to retard corrosion."
2. Limit the variables to be tested: If you change several aspects of your samples, you will not know which one actually caused the result you are observing. It is best to change only one variable across several samples to ensure the integrity of your results. "In the past I have simply prepped and then painted my iron rings with black latex paint. In this experiment, I will test the effect of using different primers prior to applying the black latex paint."
3. Establish a control: One of the samples you will test must be the control. The control is the standard by which all other samples are measured. If you are testing a change in practice, then the control is simply the old practice. "In testing the rings, a prepped but unpainted ring will be the control. We will test the past practice of prepping and painting with black latex paint. In addition, we will test four primers that will be applied over the prepped ring and then coated with the black latex."
4. Define and adhere to clear procedures: Write instructions for all steps of the process of preparing the tests, managing the samples during the time of testing, and final measuring to determine results. It is very important to standardize the handling of all samples, and to ensure that those handling the samples understand the procedures.
5. Ensure that all your samples are treated the same: Each sample should not only be handled in a standardized manner, but should be exposed to exactly the same environmental conditions and forces. "All the iron rings in the experiment will be mounted on a spruce 2"x4" base, attached in the same direction so that they cannot move, and placed on top of the workshop with the base oriented north and south. It will remain there for one year undisturbed, and will be inspected daily Monday through Friday to ensure that no debris covers any of the rings differently from the others."
6. Document your procedures, observations and results: Write all your notes and observations in a single notebook or other permanent location. Whether the time of the test is short or long, you will forget when you did something or exactly what you observed if you don't write it in your notebook immediately. "Al threw a rusty bar on the roof of the workshop on June 15th. The bar landed on the rings, and lay across the midline of ring samples 2,3, and 4, parallel to the base. We had failed to make our daily inspection, and did not discover this until July 4th when we crawled on the roof to watch the fireworks. We removed the bar, but noted rust stains on the surface of sample #4. No stains were observed on the other samples."
7. Quantify, measure, weigh, time, photograph, record: Your observations must be quantitative. Subjective observations are of little value. For example, noting that 'ring two had more rust than ring one' tells you very little. Noting that 'ring two had 45% of its painted surface area degraded by rust. Ring one had 40% of its painted surface area degraded by rust' tells you much more. A photograph of the rings side by side might tell you even more, but you should still keep a written record.
8. Write your findings in a permanent and retrievable location, and if possible publish them in a publication that is accessible to others: It is most important to establish a permanent location for all your findings. This may simply be a log or notebook dedicated to your experiments with materials tests. In this day and age some people prefer keeping their notebooks in a computer, but unless you can keep your computer in your hip pocket, I would advise against this. However, at some point you will transfer your raw data and observations from the field notebook to a more permanent location. Preferably, this would be on a computer. Ideally, you will then write the results of your tests and submit them to a publication or present them to a conference where others may benefit from your tests. If you cannot find a suitable publication, publish them in your organization's newsletter. Even if you are unwilling to share your results publicly, write them for your organization's internal purposes; you will probably not be there forever, and it will help your successor very much to find reliable notes and quantitative data on your experience.
Included as addenda following the next section of this paper are two reports describing materials test experiments that may provide additional insight into the process of materials testing. While a bit more formal that what the typical ship might produce for internal purposes, they do provide a useful model (and maybe even some useful results). One describes the effect of certain wood preservatives on epoxy adhesion, the other describes the effects of microbial activity on the strength of wood and epoxy substrates.
On the Use of Generic and Proprietary Products
Recently a colleague was touting to me the benefits of a particular low toxicity wood preservative. I very much appreciated his innovative thinking about the use of such a product, but was very much amused by his lack of innovative thinking in considering the price of the product. Indeed, when I told him that the raw material used in the product could be obtained at a cost of less than 10% of what he was paying for the proprietary product, he absolutely refused to believe me. I didn't have the heart to ask the fellow whether he paid extra for name brand prescriptions as opposed to generic drugs of the same molecular structure when he was ill. But the principal is the same.
With this group I'm probably preaching to the choir, and therefore will not carry on about this. But let me simply make the point that with a little searching, most products that we find on the shelves have generic substitutes that may work perfectly well. In other cases, packaged products are available in industrial or producers quantities at a far lower cost. With a little investigation of chemical supply houses and producers, we can often save ourselves quite a considerable sum of money.
In the instance cited above, I will protect my sources, but will simply note that the sodium borate solution that he was applying cost about $20.00 per pound, and he was applying something more than thirty pounds. Less than 100 miles from his location the same sodium borate was available from a raw materials supplier at $.50 per pound in 100 pound lots.
Return to the Preservation Conference Schedule page.