I got the following info from the SMTNet site a few months ago.  Thanks to Dave F, who also posts to Technet. 

I believe in your dimensions you mean micro-inches, not microns.  So, according to Norm's experiment, looks like you could have 50-75 insertions.

"...an experiment done by Norm Einarson. [Mr. Einarson is a very famous board design guru. His book "Bare Board PBW Design Manual" published by Printed Circuit Technology ph.617-272-0938 is a very good approach to board design for manufacture. You may be able to buy it from SMTA <http://www.smta.org> also.]

The experiment was to determine insertions from a gold plated board into a connector. Same thickness boards (0.062"), with 100 m base nickel. • 100m gold => 500 insertions • 50 m gold => 300 insertions • 30 m gold => 50 to 75 insertions • 20 m gold => 10 insertions. "

-----Original Message-----
From: Yu-Hung Shiau [mailto:[log in to unmask]]
Sent: December 11, 2001 1:21 PM
To: [log in to unmask]
Subject: [TN] Electroplated Edge connector insertion limit

Morning guys,
Customer bought our product for testing propose and would like to get a number about the normal allowed insertion times on the gold plating edge connectors. Is there anybody can give me an idea about how many insertion times for a hard-gold-plated edge connector could take before it malfunctions? We have 30 micron" gold through electroplating process over 100 micron" Nickle on all edge connectors in our products.

If possible, could somebody direct me to someplace where I could find relevant articles for my own study purpose. Many many thanks!!
YH Shiau