Metallurgical Considerations
Silicon may be the most important element to control for producing to the GGG-40.3 specification. It is a strong graphitizer, to assist in obtaining a fully ferritic matrix. It is also a ferrite strengthener. In strengthening the ferrite the impact transition temperature is increased. GGG-40.3 testing is performed at a specific temperature. In producing the grade we are not measuring the transition temperature, we are just interested in assuring the Impact Energy absorbed by the specimens meets the minimum. Since lower shelf temperatures are typically in the 4 to 7 joules range, the key is to keep the silicon low enough that the transition temperature is shifted lower than -20 degrees C.
Naturally lowering the silicon content will lower the carbon equivalent. This makes for a dilemma. In order to avoid shrink defects, carbon equivalent needs to be maintained near the eutectic. Micro shrinkage as a result of too low CE can occur in the test pieces. This will certainly lower the absorbed energy during the test. Pouring temperature now becomes a concern for shrink control.
Phosphorus must be kept low, since the phosphide eutectic lowers ductile iron toughness. This should be kept as low as possible, certainly below 0.025%. The key is raw material control.
In lowering the silicon to assist in lowering the transition temperature, yield strength suffers. Yield strength in GGG-40.3 must be maintained to the same degree as in the standard or non impact grade GGG-40. Nickel can be added to increase yield without increasing the ductile/brittle transition temperature. Care should be exercised, as too much nickel will cause nodule shape changes.
Almost all other elements, besides the ones mentioned, harm the metal in some way for the production of GGG-40.3 impact grades. Carbide stabilizers lower toughness, and many other elements will cause graphite formation or shape control problems.
Treatment and Pouring
Good inoculation is essential to produce good nodularity. The aim should be to maximize the number of Type 1 nodules. However, you should avoid raising the silicon content too much with the inoculation since this will affect the transition temperature. Solid mold inoculants are a good idea, since fade is minimized. Correct pouring temperature must be maintained. This will reduce shrink in the casting or the test bar.
Heat Treating
In producing 40.3 castings, a heat treatment is normally required. This is a full anneal which is done to completely transform any remaining non ferritic structure into ferrite.
One thing to remember is the strong influence silicon has as a graphitizing element. Since one metallurgical aim for transition temperature control is to lower the silicon content, the iron will tend to form more pearlite in the as cast condition. Then, through annealing, the pearlite can be decomposed into ferrite with the carbon migrating to existing graphite nodules or interstitial sites.
Nodule spacing becomes important. The smaller the nodule spacing, the shorter the migration path for the carbon. It should be controlled so that your heat treatment can be standardized. At the very least it should be measured, so some remedial action can be taken if the first heat treat is unsuccessful.
Micros should be taken from the impact specimens themselves until you can be certain your heat treatment cycle is producing a fully ferritic material. |
Removal of specimens from an attached bar
I think it is a good idea to specify how the samples will be taken from the bar. We usually mark the bars before having the sectioning done. The actual testing is done at an outside laboratory. We require that the laboratory maintain traceability of the three specimens. We also request that the specimens for Charpy test be returned to us. That way if there is a low result, we can examine the bar to determine why the result is low. That is why marking the bar is important.
Specifying test methods
The most important part of meeting the 40.3 specification is the actual impact test.
DIN 1693, Parts 1&2 both specify that the test specimen be machined to a DVM shape. This DVM shape is described in DIN 50015. The test coupons are 55mm long and 10mm x 10mm rectangular bars. A notch which is 3mm deep by 2mm wide is cut into the bar. The bottom of the notch is fully radiused. The uncut cross section remaining is 7mm x 10mm.
Other shapes are sometimes specified, by customers, but this U shaped notch is the standard in DIN 1693 Part 1. Good machining of the notch is important to the test.
Make sure that the lab knows what the test temperature is to be. This must be absolutely clear. Minus 20 degrees centigrade is the GGG-40.3 requirement. When we first began producing to GGG-40.3 the lab would sometimes test at the wrong temperature or test at Fahrenheit temperatures.
Test results are preferably reported in Joules. Ask the lab to report to as much precision as possible. Accuracy is of course a given requirement.
Notch Styles
Now we talked about how other notch designs are sometimes specified. The V notch is 2mm deep, comes to a point at a 45 degree angle, and the bottom has a .25mm radius. The DVM notch is U-shaped. Although the notch is 3mm deep, the sides are parallel and there is a relatively large 1mm radius at the bottom. The DVM notch is the requirement for DIN 1693, 40.3 grades.
Investigations of the specimens
After testing, the fracture surface should be examined. Look for percentage of crystalline fracture vs non-crystalline. Micros can be made of one half of the two notch pieces. This is especially important if test requirements are not met.
The future of DIN 1693, Parts 1 & 2, GGG-40.3
At this time we have been talking about 40.3, but the DIN standards of the 1970's were superseded almost two years ago. EN 1563 is the replacement specification. It was issued in August 1997. The old material designations are obsolete. There are no longer two parts to the specification, they are now combined. 32 new grades have been added and grades based on hardness are now available. They are not in wide use yet, but some inquiries have been made. Is 40.3 equal to the new standard? Maybe yes and maybe no. |
