Not every hyperbola is a rebar

Learning objectives

• Read a raw 2.7 GHz radargram: direct wave, hyperbolas, slab bottom if visible.
• Link the apex of a hyperbola to the position and depth of a target.
• Measure the effect of the dielectric constant on the calculated depth.
• Compare parallel profiles to follow a target in space.
• Understand why a single 2D profile can mislead, and what a 3D grid adds.

Prerequisites: GPR principles, notions of permittivity and propagation velocity.

The dataset

A reinforced concrete slab surveyed with a GSSI StructureScan Mini XT/LXT antenna (2,700 MHz) on a 1.20 × 0.60 m orthogonal grid with 5 cm spacing: 38 profiles.

• Profiles 001 to 013: along X (0 to 1.20 m), located at Y = 0, 0.05, …, 0.60 m.
• Profiles 014 to 038: along Y (0 to 0.60 m), located at X = 0, 0.05, …, 1.20 m.
• Origin: bottom-left corner, seen from above.
• Files: FILE__001.DZT to FILE__038.DZT. Raw data, with no gain, filtering or migration.

The area contains the rebar mesh and an electrical cable crossing it diagonally. Partial ground truth, established by electromagnetic detection.

Preparation

• Download the archive from the dataset page (free form) and copy it to the lab computers. GPRviewer reads the ZIP archive directly.
• GPRviewer runs online, in the browser: nothing to install. Just check that the computers have Internet access.
• Prepare a sketch of the grid (X and Y axes, profile numbers): students will plot their observations on it.

Session outline

1. Getting started (5 min). Open FILE__012.DZT, a profile along X located at Y = 0.55 m. Try several palettes, zoom, display the A-scan, remove the background. Identify the direct wave.
2. Reading hyperbolas (10 min). On profile 012, locate each hyperbola and note its position along the profile. Is the spacing regular? Does one hyperbola stand out by its amplitude, shape or depth?
3. Depth and permittivity (10 min). Calibrate the depth axis, then read the apex depth of each hyperbola for dielectric constants of 4, 6, 8 and 10. Compare the differences and discuss which value suits concrete.
4. Following targets across profiles (12 min). Open the neighbouring profiles 010, 011 and 013, 5 cm apart. Plot each hyperbola's position on the sketch. Repeat on a few profiles along Y, for example 016 to 020.
5. Conclusion (8 min). Join the plotted points and discuss. Then show the horizontal slice (C-scan) of the grid, as a GPRviewer Pro demonstration or a video.

Questions

1. What is the direct wave, and why does it appear at the top of every profile?
2. How many hyperbolas do you count on profile 012, and at which positions?
3. Why does a bar perpendicular to the profile produce a hyperbola rather than a single point?
4. How much does the estimated depth change between εr = 4 and εr = 10? What does this mean for coring or drilling?
5. How can the dielectric constant of concrete be determined on site, without coring?
6. On profiles 010 to 013, which hyperbolas keep the same position? Which one moves, and by how much per profile?
7. What does this tell you about the orientation of the corresponding target? Is it a rebar?
8. Why could an operator who surveyed only profile 012 have drilled into this cable? Which survey method would have prevented it?

Going further with GPRviewer Pro

• Horizontal slice (C-scan): the cable appears diagonally across the mesh, confirming the observations plotted on the sketch.
• Hyperbola fitting: the shape of the hyperbola gives the propagation velocity, hence the dielectric constant of the concrete, without coring.
• Rebar picking: spacing and cover, with statistics.
• Multi-profile 3D view: reading the grid as a volume.

Answer key

The answer key (positions and depths, annotated C-scan) is reserved for teachers and trainers. Request it from info@gprbase.com using your institutional address, stating your institution and the course concerned.

Licence and citation

Data under the CC BY-NC-SA 4.0 licence, free for teaching. Please cite the dataset in your course material: the citation is ready to copy on its page.