The following is an excerpt.

The shock wave produced by an explosion may be accompanied by fragments accelerating to immense speeds. This devastating combination is a massive challenge for those working with blast-resistant designs. The most destructive loads that exist are the topic of Benjamin Stavnar Elveli´s PhD-thesis.

Explosion test showing fragments and the shockwave.
Detonation of a vehicle borne improvised explosive device, illustrating accelerated fragments moving together with the shock wave. The propagation of the shock wave is indicated with the white arrows (Courtesy of The Norwegian Defence Estate Agency).

During combined blast and impact loading, the projectile-like fragments will often hit before the shock wave arrives. Then the chance of more significant damage increases. Buildings, cars, or other structures nearby the detonation will be subject to a loading case more severe than if either load act alone.
«This is the scariest load out there. It works the same way as shrapnel bombs», says Benjamin Stavnar Elveli, who has written a doctoral thesis on this topic.
It is known from experience that the combined effect of blast and fragment impact may be more severe than the effect of the blast or impact alone. Despite these observations, design codes do not cover such loading scenarios, and few studies are available in the open literature. Benjamin Stavnar Elveli’s PhD work in the field is a step towards better protection and safer solutions.

A DOCTORATE IN SEVERE BLAST EVENTS

It was in 2018 that the busy skateboarder, surfer, climber, and snowboarder Benjamin Stavnar Elveli was offered a doctorate in severe blast events at SFI CASA. Already, he had a master’s degree in blast loading. The task was to investigate how combined fragment impact and blast loading affect thin steel plates.
«I cannot claim to be a straight-A student during my master’s. Besides, it has always been important to spend time on my hobbies outside of working hours. So, I was unsure if I would fit the typical PhD candidate profile. However, I am glad I decided to try it».

SAFER STRUCTURES, SAFER SOCIETIES

He submitted the thesis in October. After four years of effort in solving complicated challenges, the doctorate title is within reach. The basis of his work is more than 80 small-scale blast experiments. He has used target plates of three steel types, all perforated with holes mimicking fragment impact. By combining physical blast tests with theory and mathematics, he recreates the blast loads in computer simulations. The aim is to predict and control how civil engineering structures respond to the loads.
This is science for a safer society. Understanding the load’s physics will enable the next generations of engineers to design more accurate, safer, and more sustainable structures.

Whether accidental or intentional, explosions can cause massive damage. Fragments can originate from an explosive device, such as a ruptured casing or ball bearings. Or from gravel, rocks, or debris from surrounding structures. Depending on the proximity to the detonation, fragments can strike a structure before, during, or after the blast wave arrives. If they hit before, they introduce areas of weakness, which must then withstand the incoming blast wave. It is often in these weak points that fracture is initiated. According to Benjamin Stavnar Elveli, scenarios like this are assumed to have the highest damage potential. Because the structure will already contain a defect before it can withstand the blast wave.

A COMPLEX MODELLING TASK

Detonation test showing fragment impact and a blast wave

The pressure from a blast wave can last for several milliseconds and cause deformation over a large area. Fragments travel at high speed and inflict concentrated damage. Modelling the combined effects of these threats requires that two very different, damaging loads are described within the same model. To name it a complex modelling task is an understatement.
(Courtesy of The Norwegian Defence Estate Agency).


«You will often end up with a kind of trade-off. To capture the locally reduced fracture resistance during the explosion, you must decide how accurate the descriptions of the fragment impact must be. If you do not get full control of this, you could overestimate the structure’s capacity during the blast loading».
Overestimating a component or a structure’s strength may have fatal consequences. Structural engineers must deliver reliable designs. A large part of Elveli’s work has been to investigate how accurate the models must be to ensure safe and reliable structures.

«If you do not get full control of this,
you could overestimate the structure’s
capacity during the blast loading»

FROM MASSIVE MILITARY TO LIGHTWEIGHT CIVILIAN

Historically, those working on blast-resistant design have focused on massive military concrete structures. However, new threats have emerged during the last decades. The need to protect civilian structures in urban areas has increased. And so has the interest in blast-resistant, thin-walled structures.  The attraction is due to their capability of undergoing large deformations without fracturing. This also explains Mr. Elveli’s choice of various types of steel as test material. However, establishing universal design guidelines for blast-resistant, lightweight structures takes time and effort.

READ MORE: About Blasting Steel Plates and Why Material Properties Matter

PRE-CUT DEFECTS WITH A DOWNSIDE

One common approach is to assume the fragments strike before the arrival of the blast wave. Then the loading scenario must be split into two sequential loading events. Often, such studies use structures with pre-cut defects. These mimic flaws or weaknesses from fragments before the explosion. The downside of this simplified approach is that it avoids all uncertainties related to real fragment impact.
«For instance, manufactured defects lack the small cracks that lead to fracture under blast load», according to Elveli, who used plates with idealized pre-cut defects in his master’s and first study in the PhD thesis.One common approach is to assume the fragments strike before the arrival of the blast wave. Then the loading scenario must be split into two sequential loading events. Often, such studies use structures with pre-cut defects. These mimic flaws or weaknesses from fragments before the explosion. The downside of this simplified approach is that it avoids all uncertainties related to real fragment impact.
«For instance, manufactured defects lack the small cracks that lead to fracture under blast load» says Elveli.

 

COMPARING FRACTURE STRENGTH

This leads to some essential questions and original aspects of Benjamin Elveli’s doctoral work: Do the plates with pre-cut holes behave differently from target plates with authentic holes? If yes: How does the simplified approach affect the models’ reliability on the structure’s blast resistance – and behaviour?Elveli compared the pre-cut plates with plates subjected to real ballistic impact to find the answers. As high-velocity fragment impact is very similar to ballistic impact; he fired small-arm projectiles at the target plates in SIMLab’s ballistic lab. Next, he subjected them to blast loads in the SIMLab Shock Tube Facility.

Below is an image sequence of the projectile perforation, where (a) shows high-speed photography of the ballistic impact. To the right: Images of the corresponding fracture modes for 3 different steel plates .

Projectiles perforating steel plates filmed with high-speed camera.

OVERESTIMATING STRUCTURAL CAPACITY

The pre-cut and ballistic impact holes were circular with similar diameters. But whereas the first was «clean» around the edges, the projectiles inflicted small petalling cracks and plastic deformation around the perforation hole. Under the blast loads, propagation started in these cracks. Thus, Elveli’s work demonstrates that the simplified approach may lead to models that overestimate the structure’s ability to withstand combined loads.
«Idealized defects are easier to test and to simulate. However, since they lack the deformations and damage occurring in real explosions, there is a risk of exaggerating the strength of the materials in these models».

THE PUSH FOR COMPUTER SIMULATIONS

The push for accurate computer simulations of blast-loading events is easy to understand. For many reasons, you cannot blow up real-size structures whenever you need to test blast resistance. Detonating case explosives on full-scale would, for instance, create a dense cloud of flying fragments that could destroy sensors and cameras supposed to monitor the event.

Benjamin Elveli has put a massive effort into designing controlled and reliable small-scale experiments. He has performed 110 tests, 82 of them representing the blast load physics. Cameras filming 37 000 frames per second have captured the combined loads. In the spring 2022, a test series was conducted at the Norwegian Defence Agency´s test facilities. Here he is with NDEA-researcher Ole Vestrum, preparing a test in a rig used for partially confined detonations.

READ MORE: CASA and NDEA in Close Collaboration on New Close-range Explosion Studies
READ ALSO: A Shared History. Created by Key Players.

EXTENSIVE, EXPENSIVE – AND HOPEFULLY USEFUL

«It has become an expensive doctorate, so I hope the new knowledge will be useful to more people than me», he jokes. When asked whom he hopes will benefit from his work, he points to the scientific community, both military and civilian. Creating accurate and reliable simulations is time-consuming, so the result has yet to be ready for the industry at this stage. However, the experiments generated enormous amounts of data that would interest people working in this field. Or people in the research and development departments of large companies.

READ MORE: Into the Physics of Extreme Blast-structure Interaction

«The extensive dataset is well suited to evaluate more numerical methods and develop new computational methods in the future. Also, it may be useful for engineers working with Finite element software. The complicated load cases can really enable people to test the various numerical methods in use today».

On 14 December, the 31-year-old PhD candidate defends his thesis at NTNU. The title is «Behaviour, modelling, and simulation of thin steel plates subjected to combined blast and impact loading». His supervisors are Associate professor Vegard Aune (main), and Professor Tore Børvik.