Cite this document

ASTM Suggested Citation Style, Chicago Manual of Style 17th Edition Publication Manual of the American Psychological Association, 6th Edition MLA Handbook for Writers of Research Papers 7th Edition URL

This note provides a comparison and analysis of durometer hardness Type A and Type M per ASTM D2240-15, Standard Test Method for Rubber Property – Durometer Hardness, measured on various commercial grade fluorocarbons conforming to AMS7276, Aerospace Material Specification for Rubber: Fluorocarbon (FKM) High-Temperature-Fluid Resistant Low Compression Set for Seals in Fuel Systems and Specific Engine Oil Systems. This note addresses the seemingly trivial but actually complex problem of how to translate Type A hardness range requirements measured on disks/buttons/plied platens to Type M hardness range requirements measured on O-rings of various cross sections. It confirms there is no simple correlation between Type A and Type M hardness measurements. It is shown that mapping the current Type A hardness range of 70–80 measured on buttons/disks/plied platens onto Type M hardness measurements made on O-rings will require broadening the Type M permitted range to 75, +11, −3 [72 to 86 or 79+/−7]. A tolerance range broader than the typical +/−5 is not without precedent, since the tolerance range for Type A measurements for conductive shielding gaskets per MIL-DTL-83528, General Specification for Gasketing Material, Conductive, Shielding Gasket, Electronic, Elastomer, EMI/RFI, is also +/−7. The data confirm the literature reports that Type M hardness measurements on O-rings are usually higher than Type A on disks/buttons/plied platens, but the difference decreases as the O-ring cross section increases. Though the permitted Type A hardness range for fluorocarbons conforming to AMS7276 is 70–80, actual commercial production is overwhelmingly at the high end of this range and is not centered on the midpoint of 75. The typical Type A hardness requirement range in industry as well as in AMS7276 is only 10 hardness units, but normal measurement variation consumes 45–60 % of that range; hence, obtaining simple correlations between Type A and Type M measurements is not possible because of inherent measurement variation.