How do engineers know a building system will perform when it matters most? In this episode of Building with Metal, host Kathi Miller welcomes back Bob Zabcik, Member and Owner of Z-Tech Consulting—one of our most popular guests—to explore the science behind fire and earthquake testing. Bob takes us inside today's testing laboratories to explain how building systems are evaluated under extreme conditions and why those results play such a critical role in building codes, product development, and public safety.We discuss what engineers look for during fire tests, how seismic testing has evolved, and why testing is about far more than simply passing or failing. The conversation also features an incredible look at multi-hazard testing, where researchers subjected a full-scale building to a simulated earthquake before exposing it to fire—providing valuable insight into how structures perform during real-world disasters.
Whether you're an architect, engineer, contractor, or simply fascinated by how buildings are designed to withstand nature's toughest challenges, this episode offers a behind-the-scenes look at the science that helps make our built environment safer and more resilient.
Bob Zabcik: "Strength is the amount of force it takes to break something. You hear that called rupture. Ductility is the amount of deformation a specimen can endure prior to breaking."
Bob Zabcik: "I'm almost going to accept the fact that the building is going to fail, or at least that it won't be usable when it's done. That's not really the goal. The goal is to keep the building standing and stable so the occupants have time to get out."
Bob Zabcik: "There's really no substitute for good passive fire resistance. I feel like that's where some of our best lessons are going to be learned."
[00:00:17] Cold open: engineers built a full-scale building, shook it, then set it on fire
[00:02:07] Strength versus ductility, and why standing seam roofs often beat through-fastened roofs
[00:04:37] What has changed in testing over 20 years, and the difference between R&D tests and standardized tests
[00:06:19] A typical day in a test lab, blast screens, and flying bolt heads
[00:08:39] How much of the building code comes from testing versus post-event field data
[00:10:02] Earthquakes and the fires that follow, plus active versus passive fire protection
[00:12:01] The gap nobody has measured: how much fire resistance a damaged assembly loses
[00:13:29] The shake table video, six degrees of freedom, and what the drone footage shows
[00:17:57] Newton's second law, mass, and how a response spectrum gets built and reused
[00:20:29] Where AI fits in: digesting the data and finding the response modes the test never produced
[00:24:26] Bob's takeaway: steel, ductility, and standing on the shoulders of giants
Mentioned Resources
Bob Zabcik has run test labs. He says most days in one are boring.
“It’s not uncommon to do two days of preparation, run a test over the course of an hour, and then have two more days of takedown,” he told host Kathi Miller on the Building With Metal podcast.
Then he described the other kind of day. Bolt heads popping off and embedding themselves in the insulation 25 feet up. Insulated metal panels almost exploding under load. Chunks of concrete flying around during insert testing. “So there’s a reason there are blast screens in labs.”
The test at the center of this episode is one of the big ones.
The building was 10 stories and 103 feet tall, framed in bearing-frame cold-formed steel with lightweight concrete floors. The walls were sheathed with cementitious fiberboard, XPS insulation in the cavity, and drywall inside. No cladding. Bare bones by design.
That kind of construction is currently limited to about six stories under the building code. So a 10-story version is new territory on its own. Members of the Steel Framing Industry Association funded the materials.
It went on the UC San Diego shake table, which the National Science Foundation rebuilt in 2022 for $16.3 million to give it six degrees of freedom. That means rotation as well as movement in all three directions. The table sits on pistons and wheels, with each wheel on a plate that raises and lowers, which generates the vertical motion and twisting.
Then they ran a modified version of the 1989 Loma Prieta earthquake through it. Most people know that one as the World Series earthquake. It was a magnitude 6.9, and for this test it was scaled up to a one-in-2,500-year event, which the code calls a maximum considered earthquake.
After the shaking stopped, they added wood pallets and lit the building on fire.
Zabcik draws a line between two families of fire protection. Active protection reacts to the fire. A sprinkler system is the clearest example: heat trips the heads, water comes out. Passive protection is the building itself, always there, never activated.
Here is the problem with counting on the active side after an earthquake. “Just as gas lines can rupture, the water line could rupture too. There’s no active fire protection available right after the event. All you’ve got is passive protection at that point.”
And passive systems get tested in pristine condition. They are assembled in a furnace, run through the standard, and rated. Very few of them get retested after an earthquake has worked the joints loose and popped some screws.
“Nobody really knows to what degree that takes away from the protection,” Zabcik said. He calls it a gap in understanding, and closing it was one of the main reasons for this test.
What they found: the system performed well in general, but flame and smoke did propagate through joints that had loosened. The same thing happens in buildings of any framing type. The researchers measured the joints and cracks as they took the building apart, and that data drives the next round of work.
Early in the conversation, Kathi Miller asked about a test where a product surprised him. Zabcik went straight to roofing.
Standing seam roofs often outperform through-fastened roofs, even when the through-fastened roof has a lot more fasteners in it. The reason is the clip. A standing seam clip is designed to move and distort without breaking, which lets the panel deform and take more load before anything fails.
That gets at a distinction worth knowing. “Strength is the amount of force it takes to break something. You hear that called rupture. Ductility is the amount of deformation a specimen can endure prior to breaking.” Put high strength and high ductility together, and you get toughness, which is the amount of energy required to break the thing.
A through-fastened roof is strong. It can also fail with very little visual warning. A standing seam roof breathes and moves, and you can watch all of it happen in a test.
Miller’s reaction was honest: they have talked about the price difference and the pros and cons of the two systems many times, and never once about this.
Seismic design works differently from wind and snow design, and Zabcik was blunt about it.
“I’m almost going to accept the fact that the building is going to fail, or at least that it won’t be usable when it’s done. That’s not really the goal. The goal is to keep the building standing and stable so the occupants have time to get out.”
Time is the whole currency. The building has to stay stable, and people have to be able to get out. Fire cuts into that window, which is why passive resistance carries so much weight in seismic areas. “There’s really no substitute for good passive fire resistance. I feel like that’s where some of our best lessons are going to be learned.”
The physics behind it is high school Newton. Force equals mass times acceleration. The ground is going to accelerate however it accelerates, and a designer has no say in that. Mass is the part you can control, which is part of why lightweight steel framing does well here.
Feed all the motion data into a computer, and you get a response spectrum, which describes the deflection and force to expect for a given type of movement, acceleration, and building period. Once you have a response spectrum for a system, you can apply it to different ground acceleration maps and analyze that building under earthquakes that were never run on the table.
Miller asked whether AI has a role in a testing world producing hundreds of thousands of instrument readings. Zabcik said there is no way a human can work through that volume in a practical way.
The first job is digestion: pointing analysts toward specific behaviors buried in the response spectra. The second one interests him more.
During an event, a building falls into response modes, periods where it moves in a pattern rather than randomly. Ground acceleration changes, the pattern breaks, and a new mode takes over. There are close to an infinite number of possible modes, and a given earthquake will only produce a handful.
“It could say, ‘It missed this response mode, but if the acceleration had been different, that response mode could have occurred and affected the design of the building.’”
He also described something already in progress that he called mind-blowing: running a shake table test on the West Coast while a university on the East Coast loads an identical column or beam sample in real time, matched to the same ground accelerations at the same time steps.
Asked for one takeaway, Zabcik gave the tongue-in-cheek answer first. “I would say steel is the best building material there is.” Miller: “I didn’t even pay you for that.”
Then he gave the real one, by way of Isaac Newton: “If I have seen further, it is by standing on the shoulders of Giants.”
His point was about the people behind the codes. Hundreds of thousands of engineers, architects, and building professionals got the industry to where it is, and almost none of their names are known. Codes get criticized for being slow and frustrating, and engineers take some of the blame. The problems are hard, and they take a while to solve.
“But we all care about public safety. We’ve taken oaths. We work together. Any one of us depends on the work of hundreds or thousands of people who came before us.”
He closed by pointing to those doing the work now, including younger professors and young women running these projects. He mentioned the engineering community getting more diverse and said he is excited about it. Miller noted that one of her recent guests was Rita Brown, president of the National Association of Women in Construction, who would be glad to hear it.
McElroy Metal manufactures metal roofing and siding products for residential, commercial, agricultural, and industrial applications. Headquartered in Bossier City, Louisiana, with plants and service centers across the United States, McElroy supplies contractors, architects, designers, and building owners with metal panel systems, accessories, and the engineering expertise to use them well.