Welds at heart of Quad building collapse investigation
The integrity of the construction welds of Quad/Graphics Inc.’s Lomira plant are at the heart of the controversy over who ultimately was responsible for the collapse of the building.
A robotic welding machine operated by Livonia, Mich.-based Rack Structures Inc., performed many of the welds used in construction of the Lomira plant.
The robotic welds by Rack Structures were miscalibrated, according to independent welding experts hired to monitor the fabrication process.
Rack Structures was a subcontractor hired by HK Systems, New Berlin, the prime contractor for the Lomira building.
Inspectors found that welders at Rack Structures and at a Wausau fabrication shop were not properly credentialed to perform the welding tasks they were charged with.
E-mail correspondence during the construction of the Lomira plant from Bruce Lammi of Graef, Anhalt, Schloemer & Associates (GASAI), Milwaukee, indicated concerns over the welds.
Scott Denzin of Denzin Engineering, Wausau, said his conscience is clear regarding the quality of his welding work for the Lomira plant.
"Bruce Lammi said my welding was good," Denzin told Small Business Times.
Denzin was the only subcontractor unafraid to violate Quad/Graphics’ request that the subcontractors refrain from discussing the collapse of the building. "Quad themselves was always happy with my work."
Denzin acknowledged that there were "issues with certification" at his company.
A June 19, 2002, report from the Milwaukee-based inspection and testing firm NDT Specialists Inc. stressed that none of the welders involved in the Lomira project were properly certified.
The Wisconsin Department of Commerce’s notice of violations for the Lomira plant made similar observations in its report on the collapse.
Denzin attributed the lack of proper certification to confusion about the specified materials and processes in the project.
"I didn’t design it," Denzin said. "I built the racks with customer-supplied materials and components. I did not supply the parts. They had referenced the wrong welding spec in the project, and it never got settled."
Denzin said he is aware he may ultimately be a target for a lawsuit over the collapse, but he said he will be prepared.
"I have 1,700 to 1,800 photos of good welds," Denzin said. "We were repairing the other welds (performed by Rack Structures) as we were making the parts."
Documentation arising from the DOC investigation mentions some bad welds at Denzin’s facility.
However, after reviewing material related to the investigation, a University of Wisconsin in Madison welding expert pointed a more damning finger at the faulty robotic welding system at Rack Structures.
Upon viewing reports and photos of the faulty welds, Sindo Kou, chair of the department of Materials Science and Engineering at UW-Madison, said he could not understand how so many bad welds could happen in the construction of the Lomira plant.
Kou noted a large number of welds that exhibited a lack of fusion and insufficient penetration. When he was told that a robotic welding machine performed many of the welds, he said it all made sense.
"Robotic welding – I am not so surprised," Kou said. "If it was a person, how can he ignore a weld that has no fusion at all? Just a little bit of deviation can cause a problem if you rely on robots. If you are off one-sixteenth of an inch, or if your electrode is off, it can cause problems. I know the robot itself is pretty precise. But I don’t know if the structure itself is."
Insufficient penetration involves a failure of weld fusion to reach deeply enough into a joint. Lack of fusion implies that there is an incomplete attachment between the weld and the substrate metal.
Other problems with the welding process resulted from mistakes in the plans, according to the DOC investigation.
In correspondence with the DOC, Robert Chi, whose one-man engineering firm, EGE Inc., of suburban Chicago, served as an independent contractor to Rack Structures, acknowledged that a drafting error was made, resulting in the incorrect type of weld on the load arm of the rack system.
"The load arm should be flare-bevel-groove welded to the column," Chi wrote. "The fillet weld shown (in the plans) was a drafting error."
March 21, 2003 Small Business Times, Milwaukee
